Epitome mapping of E6 and E7 proteins from high-risk oncogenic HPV types 16, 18 and 45 CURRENT

Background It is crucial to reveal entire epitomes of more target homologous proteins and specificity of each mapped B cell epitope (BCE) within them for the development of high-risk (hr-) human papillomavirus (HPV) type-specific diagnostic reagents. Methods Recombinant E6 and -E7 oncoproteins from HPV16/18/45 were immunized mice to prepare Rabbit antisera. Overlapping 16mer/8mer-peptides for two rounds of antigenic peptide and fine BCE motif mapping were expressed as GST188 fusion proteins. Fine BCEs were delineated by Western blot and sequence alignment. Results In this work, we decoded six epitomes of E6 and E7 oncoproteins from three HPV types 16, 18 and 45 that are the most common hr-HPVs in cervical cancer patients worldwide, in which total 35 fine BCEs (8, 6 and 4 for E6; 7, 6 and 4 for E7) were mapped using rabbit antisera to respective recombinant proteins. The specificity of each mapped BCE among 20 defined or possible hr-HPVs was delineated by sequence alignment based on BCE minimal motif. According to similarities of immune responses to E6/E7 existed among rabbit and humans, 7 human-recognizing (HR) BCE motifs in HR-peptides of HPV16/18-E6 and E7 proteins were delimitated by comparing with corresponding rabbit-recognizing BCEs. Also, the unique BCE distribution within three delineated E7 epitomes was confirmed, in which almost mapped BCEs were clustered at the first half of the molecules, suggesting that it may be a common characteristic of hr-HPV E7 proteins. Conclusions The results would form the basis for identifying HR-BCEs and developing serodiagnostic reagents used in HPV-based cervical cancer screening and HPV-positive women managing. on E6 of HPV18 was identified, and the unique BCE distribution characteristics on E7 proteins of HPV16/18/45 were confirmed. Seven HR-BCE motifs on E6 and E7 of HPV16/18 in known HR-peptides were delimitated according to the similarity of immune responses among rabbit and humans. The results will facilitate the delimitation of HR-BCE motifs and the development of multi-BCE peptide diagnostic reagents and/or BCE peptide chips for the HPV type-specific detection, and lead to a better understanding of similarity and difference of immune

electrophoresis by rearranging the gel-carrying plate when the bromophenol blue band reached 1-1.5 cm from the bottom of the gel, and 8 fractions of 3-4 mL each were collected after the bromophenol blue had run out of the gel into a collecting trough between the gel and a dialysis membrane used to isolate the upper tank buffer; vi) from each collected fraction, 0.2 mL was taken and the r-protein was precipitated by sequentially adding trichloroacetic acid and acetone; vii) the sample was dissolved in 20 µL 4 × sample loading buffer, analyzed and quantified (together with 2 µg BSA standard) by SDS-PAGE; viii) finally, the identified fractions were pooled and the precipitated r-proteins were used as immunogens.

Biosynthesis of 8/16mer-peptides
The designed overlapping 16mer/8mer-peptides for two rounds of antigenic peptide and fine BCE motif mapping of E6 and E7, which all had an overlap of 8 and 7 aa residues respectively, were expressed as GST188 fusion proteins as described earlier [36]. Briefly, the synthesized annealed DNA fragments encoding each 8/16mer-peptide, incorporating BamH I and TAA-Sal I cohesive end on their 5′and 3′ends, were inserted into the polyclonal region downstream of GST188 gene in pXXGST-3 plasmid; BL21 (DE3) pLysS E. coli cells were transformed with r-plasmids; respective two or three clones grown on the LB-amp plates were picked-up to conduct thermo-inducible expression and the harvested cell pellet was first used to run 15% SDS-PAGE to determine respective r-clones according to the specifically expressed short peptide fusion proteins with about 23 and 24 kDa, since there is no any visible specific band in the weak antigenic area of proteins from non-recombinant clone. Finally, the r-clones expressing 8/16mer-peptides as fusion protein were confirmed by DNA sequencing of inserted gene fragments. All r-clones contained each fusion peptide were stored at -20°C.

ELISA
Serum antibody titers were determined by ELISA using respective r-protein as the antigen. ELISA plates (96-well; Sigma-Aldrich Inc., USA) were coated with 400 ng of r-E6, -E7 or -E6E7 protein/well.
After overnight incubation at 4 °C, excess liquid was removed and non-specific sites were blocked by the addition of 200 µL of 3% BSA in PBS (pH 7.4) for 1 h at RT. Following blocking, plates were washed three times in PBST (PBS supplemented with 0.05% Tween-20), and then serially diluted rabbit antisera in PBS were added and incubated for 2 h at RT. After washing three times with PBST, goat anti-rabbit IgG conjugated to HRP at an optimized dilution of 1:1000 in PBS was added to each well and incubated for 1 h at RT and washed as described above. Plates were incubated with 0.4 mg/mL orthophenylenediamine in 50 mM citrate phosphate buffer, pH 5.2 with 0.015% (v/v) H 2 O 2 as the substrate to develop color. The reaction was stopped with 50 µL of 4N H 2 SO 4 and the absorbance was read at 490 nm with 630 nm as reference filter using an ELX 800 Universal Micorplate Reader (Bio-TEK Instruments, Inc., USA).

SDS-PAGE and Western blotting
Cell pellets obtained from 3 mL culture of expressed r-proteins and peptide fusion proteins were boiled in 400 µL of 1 × sample loading buffer (50 mM Tris-HCl, 2% SDS, 0.1% bromophenol blue, 10% glycerol and 100 mM β-mercaptoethanol) for 5 min and proteins were separated by 15% SDS-PAGE under reducing conditions. Two gels running same samples were either stained with Coomassie brilliant blue G-250 for analyzing the bands of fusion proteins or performed for Western blotting by electrotransferring the proteins onto 0.2 µm nitrocellulose membrane. The complete transfer of proteins was ensured by staining the nitrocellulose membrane with 0.1% (w/v) Ponceau S.
Nitrocellulose membrane was subsequently processed for Western blotting using rabbit pAbs against r-E6, -E7 or -E6E7 fusion protein (1:300 dilution in PBS containing 0.05% Tween 20 and 1% skim milk powder) or preimmune rabbit serum. Specific antigen-antibody reaction on the membrane was visualized by using goat anti-rabbit IgG or goat anti-mouse IgG (only for identification of expressed HPV18-E6) conjugated to HRP at 1:2000-dilution. The blot was developed by using the ECL plus Western blot detection reagents.

Sequence alignment of homologous HPVs
The aligned E6 and E7 sequences of prototypes from hr-HPV16 to hr-HPV82 within Table 1

Expression, purification and immunization of E6 and E7 proteins
To obtain rabbit sera against respective immunogens for decoding IgG-epitomes of E6 and E7 proteins from HPV16/18/45, expression of all target proteins was carried out employing E. coli expression system. Since E6 and E7 proteins of HPV16 and 45 failed to be solely expressed in E. coli with many various prokaryotic expression plasmids, they were finally expressed in the form of fusion protein in E. coli ( Fig. 1A and 1C), respectively, using the synthesized fusion genes encoding E6E7 proteins of HPV16 and 45, which were inserted into pET28a plasmid. The expression of HPV18-E6 and -E7 proteins was achieved using pBV221 and pRSET-A plasmids, respectively (Fig. 1B). All r-E6E7 fusion proteins of HPV16/45, r-E6 and -E7 proteins of HPV18 were purified with the PAGE-preparation method, and identified with mAb to His6-tag or mAb C1P5, respectively ( Fig. 1A-C). After obtaining four r-proteins with the purity higher than 95%, they were used as immunogens to immunize rabbits, which were emulsified in complete Freund's and incomplete Freund's adjuvants. The collected antisera were used to determine the Ab titers by ELISA and the results showed that all r-proteins elicited high levels of antigen-specific Ab titers ( Supplementary Fig. 1).

E6 and E7 epitomes of HPV16
HPV16 is the first to be found that it is associated with invasive cervical cancer [39, 40] and the most prevalent HPV type in the world. HPV E6 and E7 oncoproteins can regulate the cell cycle through reacting with p53 and pRB to contribute the progression of invasive cervical cancer [9-12]. However, only limited information were available about the humoral immune responses to E6 and E7 proteins, in particular linear BCEs, although for them there were several studies on BCE mapping with sera from cervical cancer patients with HPV16-positive, murine pAbs and mAbs [24][25][26][27]. Thus, it is still necessary to delineate all BCEs and type-specific BCEs in HPV16-E6 and E7.
To reveal non-conformational IgG-epitome of HPV16-E6 with 158 aa, in the first round of antigenic peptide mapping, 19 of 16mer-peptides (a1 to a19, a19 is 14mer) with an overlap of 8 aa covering the full-length sequence of E6 were expressed as r-protein fused with GST188 carrier in E. coli using the pXXGST-3 plasmid. For the second round of fine BCE identification, 75 octapeptides (a20 to a94) with an overlap of 7 aa corresponding all reactive 16/14mer-peptides shown in Western blotting were expressed in E. coli. In epitome mapping of HPV16-E7 protein, 12 of overlapping 16mer-peptides (b1 to b12, b12 is 10mer) and 56 overlapping 8mer-peptides (b13-b68) corresponding all reactive 16merpeptides were employed.
The position of 15 delineated BCEs on two predicted three-dimensional (3D) structures of HPV16-E6 and E7 proteins has been shown in Fig. 3A and 3D, wherein 3D structures of E6 and E7 are predicted according to the previously described method [41]. Of all mapped BCEs on these two proteins, 2 BCEs (16/E6-5 and 16/E6-6) on E6 and 1 BCE (16/E7-7) on E7 were located in predicted α-helical domains, respectively. The linear schematic diagrams of decoded E6 and E7-epitomes in the predicated 3D structures of E6 and E7 showed two obvious features: i) the distribution pattern of eight BCEs on E6 presenting disperse state; and ii) except for a BCE 16/E7-7 at the C-terminal of E7, the rest of 6 BCEs being situated in the first half of the protein, like those (7/8) of the HPV58-E7 epitome [30].

E6 and E7 epitomes of HPV18
HPV 18 is associated predominantly with adenocarcinomas and adenosquamous carcinomas compared with squamous cell carcinomas [42,43] and is the second most common type in invasive cervical cancer worldwide and in China [5,42]. It is essential to reveal all BCEs on HPV18-E6 and E7, in particular the type-specific BCEs, for developing specifically diagnostic reagent of HPV18. Although there have been studies on epitope mapping of HPV18-E6 and E7 using different methods including phage display library and/or peptide pin screening and sera from cervical cancer patients with HPV18positive, rabbit pAbs and murine mAbs [27-29], the so-called BCEs revealed are longer antigenic peptides, and information about their fine and type-specific BCEs is still lacking.
For epitome mapping of HPV18-E6, 19 of overlapping 16mer-peptides (c1 to c19, c19 is 14mer) covering the full-length sequence of HPV18-E6 with 158 aa were used in the first round of BCE mapping. In the second round of fine BCE identification, 7 set of 57 octapeptides (c20 to c76) corresponding 8 mapped antigenic 16/14mer-peptides were used. As shown in Fig Table 3), which were named as 18/E6-1 to 18/E6-6. Because of failing to find any reactive band within a set of overlapping octapeptides (c45-c53) corresponding reactive peptide c15, the nonapeptide BCE of 18/E6-4 was identified by using another group of 10mer-peptides (c54-c61). The 18/E6-6 motif in reactive c71 located at the end of Pc19 was identified using five extra octapeptides (c72-c76), wherein the respective aa residue from N-terminal of c71 was sequentially substituted with A.

E6 and E7 epitomes of HPV45
The HPV45 has been determined as the third most prevalent hr-HPV type involved in invasive cervical carcinoma, and found in 4-9% of cervical cancers worldwide [3,5], which is closely related to HPV18 and belongs to HPV species 7 in the genus alpha-papillomavirus [44]. Unlike HPV16 and 18 that have been studied intensively, little is known about the biological properties of HPV45, but it has been reported that HPV45-E7 protein can be expressed in cervical cancer biopsies [45,46], and is a transforming protein that binds to pRB and induce pRB-degradation and anchorage-independent cell cycle progression [47]. Furthermore, to our knowledge, the study on BCE mapping of HPV45-E6 and E7 proteins has not yet been carried out up to now.
As shown in Fig. 5B, the consecutively reactive bands among e21-e25, e34-e37 and e43-e47 and a single reactive band of e28 in 3 sets of octapeptides were observed, and thus 4 fine BCEs ( 6 DPKQ 9 , 9 QRPYKLPD 16 , 80 YYSNS 84 and 113 NPAE 116 ) were identified according to each common sequence within reactive octapeptides (Supplementary Table 5 proteins are shown in Fig. 3C and 3F. Of mapped 8 BCEs on E6 and E7, only one (45/E6-3) was located in a β-sheet domain (Fig. 3C). Their linear schematic diagrams of mapped BCEs in E6 and E7epitiomes has been shown in the predicated 3D structures of E6 and E7. There is no distinct BCEs distribution property for E6 (Fig. 3C), and the epitome of HPV45-E7 showed the same feature of BCE distribution pattern with HPV16/18/58-E7, in which all four mapped BCEs were clustered at the Nterminal (Fig. 3F).

Specificity of mapped BCEs among hr-HPV homologous proteins
Since it has known that early E6 and E7 oncoproteins are consistently expressed in the HPV life cycle [8,9] and they are associated with cervical cancer during hr-HPV infection [10-12], the goal of most E6/E7 epitope mapping studies mainly was to identify the antibody-reactive peptides using sera from cervical cancer patients with HPV16/18-positive and rabbit/murine pAbs [23,27,29]. Obviously, it is imperative to map all type-specific BCEs on E6 and E7 of each hr-HPV among homologous proteins for the development of highly specific and sensitive detection reagents.

Delimitation of HR-BCE motifs within known HR-antigenic peptides
Based on the similarities of Ab responses to HPV18-E6 and E7 between rabbit and human [29], we investigate the cross-reactivity of mapped RR-BCEs with human sera to HPV16/18 by comparing sequences of mapped RR-BCEs with HR-peptides. Specifically, the RR-BCEs recognized by human serum can be determined according to whether they are present within corresponding HR-peptides identified by sera from patients with HPV16 and 18-positive. The results showed that 9 in 27 mapped RR-BCEs on E6 and E7 of HPV16/18 were determined to be recognized by human sera to HPV16/18 based on their presence in 8 known HR-peptides. Of them, 5 RR-BCEs including 16E6-5, 16E7-1, -2, -3,

Discussion
Using antigenic peptides to replace intact r-proteins in serological test of HPV infection is a great progress, but an ideal peptide reagent used in viral antibody detection should be a shorter peptide pool based on BCE motif and/or combines as many type-specific BCEs as possible in a r-peptide in order to improve specificity and/or sensitivity of serological diagnosis. Clearly, the key to achieve this goal is that entire IgG-epitomes of several target viral proteins and all type-specific BCEs in them can be revealed. Now, it is achievable since all linear BCEs present in a protein and type-specific BCEs in them has been able to be delineated using rabbit/murine pAbs using the BSP method specific for BCE mapping [35,36], which has been used in many studies of mapping BCEs on different viral protein using rabbit pAbs and BSPs [30, 48-52]. Our previous study has revealed three IgG-epitomes of E6, E7 and L1 proteins from HPV58 virus [30]. In this study, we decoded two sets of six IgG-epitomes of HPV16/18/45-E6 and E7 proteins using this BSP method and all type-specific BCEs within them.
The decoded homologous epitomes of E6 or E7 showed obvious differences for the number of mapped RR-BCEs (Fig. 3) and defined type-specific BCEs (Table 1) in respective corresponding proteins, even in two proteins with high aa homology, for example, E6 and E7 proteins from closely related alpha viruses HPV18 and 45 share 81% and 78% aa homology, respectively [53]. These differences, including the irregular or regular distribution characteristics of BCEs in respective homologous epitomes, indicated the necessity to carry out epitomics study on two sets of homologous proteins from HPV16/18/45. HPV18-E6 epitome mapping with rabbit pAbs shows that 6 fine BCEs all are in four longer antigenic peptides mapped in the previous study [29], suggesting the credibility of mapped HPV18-E6 epitome. Moreover, the 18/E6-3 BCE ( 86 YGDTLE 91 ) of HPV18-E6 was almost identical to the E6-2 BCE ( 84 YGDTL 88 ) of HPV58-E6 protein [30], indicating that the rabbit immune system could recognize the same antigenic site in various proteins. Interestingly, rabbits could recognize not only the similar antigenic region ( 91 YGTTLE 96 ) with one residue mutation in HPV16-E6, but also the mapped 16/E6-5 motif ( 90 LYGTTLEQ 97 ) with two more residues on its N and C-terminals compared with 18/E6-3 motif.
To our knowledge, the BCE minimal motifs mapped using mAbs and chemically synthesized peptides (CSPs) together with ELISA were 4mer-6mer peptides [24, 28, 29, 54], whereas the BCE motifs mapped using pAbs and BSPs together with Western blotting can be the shortest 3mer-peptides [30, 55] and the longest 9mer-peptide (18/E6-4) mapped in this study, even longer 10mer-peptide [52]. In addition to contributing to reveal complete BCEs on a protein using animal sera and their specificity among homologous proteins, the BCE motif identification has two other implications: i) helping to the design of multi-BCE peptide diagnostic reagents and/or vaccines, which would improve the accuracy and sensitivity of serological diagnosis by combining many specific BCEs as possible, and prevent potential harmful antibody cross-reactivity to other key proteins in human body, that is, it should avoid using the mapped 3mer and 4mer-peptide BCEs if they are used as candidates of preventive vaccine; and ii) conducing to understand the possible regular differences of BCE motifs located the same site that are mapped by various species pAbs and/or mAbs, for instance, the PTRR and ELRHY (RELRHY) mapped in this work has also one residue difference compared with MR-BCE (ELRHY) [36] identified by HPV18-E6 mAb C1P5. The similar result was also found between RR and HR-BCEs such as 16/E6-5 and h-16E6-1. In short, it should be a feasible and reliable method to determine HR-BCE motifs in known HR-peptides according to the mapped RR/MR-BCE motifs.
Mapping of BCE minimal motif is very important, since it is the basis of decoding complete IgGepitome of a target protein and then revealing specificity or conservativeness of each mapped BCE among homologous proteins efficiently by sequence alignment [30]. The antibody-recognizing BCE motif is one of major features to distinguish different antibodies as well, which obviously conduces to determine whether two or more mAbs with various IgG isotypes are identical in the function of binding target protein. For example, two mAbs E6-18-1 (IgG3) and E6-18-2 (IgG1) derived from HPV18-E6 as well as mAbs 8C6 and 1B12 produced by M2 protein of influenza A virus all can be classified as the same antibodies, since they recognize the same tatrapeptide PTRR or pentapeptide EPTIR, respectively [29, 57]. More importantly, it helps in overcoming reproducibility crisis or disorder present in antibody application that was mentioned a few years ago [58] for knowing BCE motif of each antibody, as currently users can conveniently employ BSP method to verify the obtained antibody quality of different batches or whether it is a target antibody needed if researchers or suppliers can provide each antibody-recognizing BCE motif, but which has not been included in the disorder solution as a required parameter [58,59]. In short, the meaning of fine BCE mapping should be emphasized in the fields related to BCE mapping and antibody research henceforth.
In conclusion, six complete IgG-epitomes of HPV16/18/45-E6 and E7 proteins were decoded, and 17 of mapped 35 BCEs on E6 and E7 proteins were determined as possible type-specific BCEs among 20 compared HPV types. A nonapeptide BCE on E6 of HPV18 was identified, and the unique BCE distribution characteristics on E7 proteins of HPV16/18/45 were confirmed. Seven HR-BCE motifs on E6 and E7 of HPV16/18 in known HR-peptides were delimitated according to the similarity of immune responses among rabbit and humans. The results will facilitate the delimitation of HR-BCE motifs and the development of multi-BCE peptide diagnostic reagents and/or BCE peptide chips for the HPV typespecific detection, and lead to a better understanding of similarity and difference of immune 20 responses to the same protein among different species and epitope biology.   100% conserved BCEs and BCEs that have an aa mutation but exist in at least two HPVs, or together with a conserved BCE. "− " means no the same or similar sequences.  corresponding 10 and 7 reactive 16mer-peptides of E6 and E7. The number above each membrane represents overlapping 16mer a1 to a19 for E6 and b1-b12 for E7 as well as 8mer a20-a94 for E6 and b13-b68 for E7.