The complete amino acid sequence of Pla a 2 was acquired from the Nucleotide database of the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/nucleotide/), accession number Q6H9K0.

A total of 10 adult individuals comprised two groups: i) Five allergic subjects (aged between 20 and 43; 2 males and 3 females; recruited in May 2016; diagnosis was established based on clinical symptoms associated with allergic rhinitis during the pollination season, a positive skin prick test result and a seropositive IgE test to P. acerifolia pollen extract; and ii) five healthy controls (age, 19–45; 2 males and 3 females; recruited in May of 2016). Written informed consent for the use of blood samples was obtained from the fingertips of all participants prior to study entry according to The Declaration of Helsinki. The study protocol was approved by the ethics committee of the First Affiliated Hospital of Nanjing Medical University (Nanjing, China).

The nucleotide sequence corresponding to the mature Pla a 2 allergen (signal peptides were removed, which cannot be excised by cell recognition and they may affect the proper folding of proteins.) was synthesized by GenScript (Nanjing, China) and was sub-cloned into a pET28a vector (Novagen, Madison, WI, USA) using BamHI and XhoI sites, and was verified by Sanger (20). Specifically, the full-length Pla a 2 gene was amplified by reverse transcription-quantitative polymerase chain reaction with a pair of specific primers and the PCR product was cloned into the pET28a vector using Tag enzymes (Takara Biotechnology Co., Ltd., Tokyo, Japan). The recombinant pET28a-Pla a 2 plasmid was transformed into the ArcticExpress™ (DE3) RP Escherichia coli host strain (21,22). Firstly, 1 µl pET28a-Pla a 2 plasmid was transformed into 100 µl ArcticExpress™ (DE3) Escherichia coli cells and placed on ice for 30 min and then heated in a water bath at 42°C for 45 sec, followed by an ice bath for 2 min. A total of 500 µl LB medium without antibiotics was added to the tube with shaking at 200 rpm for 1 h at 37°C. Subsequently, 200 µl transformed cells were added to the LB agarose solid state medium and placed at 37°C for 1 h. Finally, the tube was cultured overnight at 37°C.

The positive clones were cultivated overnight in 3 ml lysogeny-broth kanamycin (concentration 0.05 g/ml) at 37°C until an absorbance of 0.6–0.8 at 600 nm was attained. The culture was induced with 1 mM isopropyl-b-D-thiogalactopyranoside and harvested following incubation for 4 h at 37°C. Uninduced culture (subject to the same conditions without IPTG) was used as a control. Expression of recombinant Pla a 2 was confirmed using 12% SDS-PAGE. The cell mass from 200 ml induced culture was resuspended in 20 mM Tris-HCl buffer (pH 8.0). The cell suspension was sonicated at 40 kHz in 10 cycles of 4 sec pulse on and 8 sec pulse off at 4°C, and was subsequently centrifuged at 10,000 × g for 30 min. Analysis revealed that recombinant Pla a 2 was mainly located in the supernatant (Fig. 1); purification was conducted via Nickel affinity chromatography (GenScript) (23,24). The washing buffer contained 100 mM NaH₂PO₄, 20 mM Tris-HCl and 100 mM imidazole, pH 8.0; eluting buffer contained 100 mM NaH₂PO₄, 20 mM Tris-HCl and 250 mM imidazole, pH 8.0 (25).

Purified Pla a 2 (5 µg) protein was separated by 12% SDS-PAGE. The proteins were then transferred to polyvinylidene difluoride (PVDF) membranes and were blocked with 5% skim milk at room temperature for 2 h. Subsequently, PVDF membranes were incubated with a serum mixture (1:40 in PBS) from six patients with P. acerifolia pollen allergy, overnight at 4°C. The serum mixture served as the primary antibody. Subsequently, IgE-allergen complexes were detected using horseradish peroxidase-conjugated goat anti-human IgE monoclonal antibody (cat no. A9667; diluted 1:3,000; Sigma-Aldrich, Merck KGaA, Darmstadt, Germany) at room temperature; an ImageQuant LAS 4000 mini detection system (GE Healthcare, Chicago, IL, USA) was employed for the detection of these complexes. An enhanced chemiluminescence substrate including luminol/enhancer solution and peroxide solution was used (1:1 mixed; Bio-Rad Laboratories, Inc., Hercules, CA, USA). Serum mixture from six healthy individuals was applied as a negative control and the experiment was repeated three times.

SWISS-MODEL Repository is a database of three-dimensional (3D) protein structure models generated by the SWISS-MODEL homology modeling pipeline (26). The homologous templates suitable for Pla a 2 were selected using the SWISS-MODEL server (http://swissmodel.expasy.org/) based on its complete amino acid sequence (27). The most appropriate template was retrieved from results of previous analyses and was utilized for homology modeling.

Three immunoinformatics tools, including DNAStar protean system (https://www.dnastar.com/t-allservices.aspx), BepiPred 1.0 server (http://www.cbs.dtu.dk/services/BepiPred/) and Bioinformatics Predicted Antigenic Peptides (BPAP) system (http://imed.med.ucm.es/Tools/antigenic.pl) were used to predict the B cell-associated epitopes of Pla a 2 (28). The results obtained from the three immunoinformatics tools were combined to produce the most informative results (29). Four properties (hydrophilicity, flexibility, accessibility and antigenicity) of the amino acid sequence were selected as parameters for epitope prediction in the DNAStar protean system (30). The BepiPred 1.0 server and the BPAP system required the amino acid sequence of Pla a 2.

T cell epitope prediction was based on the identification of bound peptide fragments on major histocompatibility complex (MHC) complexes. The binding significance of each peptide to the given MHC molecule is based on the estimated strength of binding exhibited by a predicted nested core peptide at a set threshold level. NetMHCII-2.2 (http://www.cbs.dtu.dk/services/NetMHCII/) (31) was employed to investigate human leukocyte antigen (HLA)-DQ alleles. NetMHCIIpan-3.0 (http://www.cbs.dtu.dk/services/NetMHCIIpan/) was applied to predict HLA-DR-based T cell epitopes (32). In the present study, HLA-DQA10101-DQB10501, HLADQA-10501-DQB10201, HLA-DQA10501-DQB10301 and HLA-DQA10102-DQB10602 were used to predict HLA-DQ-based T cell epitopes. Combining the four results indicated the most informative HLA-DQ-associated T cell epitope result. Providing three epitopes were revealed, the consensus would suggest an epitope. This method was also applied to HLA-DR-based T cell epitope prediction. HLA-DRB-10101, HLA-DRB30101, HLA-DRB40101 and HLA-DRB50101 were used to predict HLA-DR-based T cell epitopes. The most informative results were obtained by combining the results of the HLA-DQ- and HLA-DR-based T cell epitope predictions. B and T cell epitopes identified by computational tools were mapped into a linear sequence and onto a 3D model of Pla a 2 to determine their position (33).

P. acerifolia pollen Pla a 2 was subcloned into a pET28a vector and transformed into BL21 (DE3) E. coli host strain. It was demonstrated that Pla a 2 was expressed mainly in the supernatant fraction (Fig. 1). The dissolved inclusion body of Pla a 2 was purified via Nickel affinity chromatography. The purity of the purified Pla a 2 was identified by SDS-PAGE. A single band with an apparent molecular weight of ~25 kDa was observed (Fig. 1).

The ability of Pla a 2 to bind IgE within the sera of patients associated with P. acerifolia pollen allergy was determined by western blotting. Pla a 2 demonstrated positive IgE reactivity to mixed serum of patients associated with P. acerifolia pollen allergy; reactivity to healthy controls was not observed, as presented in Fig. 2.

Protein Data Bank accession no. 4c2I revealed a marked sequence homology (25.22%) to the Pla a 2 allergen and was used for homology modeling. The predicted B and T cell-associated epitopes were superimposed on the surface of the Pla a 2 allergen as presented in Fig. 3.

The probability of epitope formation within the Pla a 2 sequence is indicated by antigenic index. Regions of high hydrophobicity are also associated with epitope identification. In addition, surface accessibility and fragment flexibility are important features for antigenic epitope prediction. From these sequence properties, the final predicted epitope regions of Pla a 2 were 15–24, 60–66, 78–86, 105–111, 117–125, 232–240, 260–268, 298–306 and 315–322 (Table I) using DNAstar. The predicted results of BepiPred 1.0 server were 2–9, 15–24, 36–46, 58–66, 74–87, 100–126, 170–179, 202–213, 219–226, 233–240, 260–273, 294–306, 317–323 and 350–373. Additionally, the predicted results of BPAP system were 5–12, 23–30, 32–38, 40–47, 67–80, 88–100, 133–144, 153–162, 164–172, 180–189, 190–207, 210–220, 224–230, 244–250, 283–301, 303–309, 322–335, 356–365 and 368–374. Potential B cell epitopes of Pla a 2 were then selected from the results of these three tools. The most informative result of the three immunoinformatics analyses led to the prediction of eight peptides (15–24, 60–66, 78–86, 109–124, 232–240, 260–269, 298–306 and 315–322), which are presented in Table II.

NetMHCII-2.2 and NetMHCIIpan-3.0 were used to identify the T cell epitopes of Pla a 2. For HLA-DQ alleles, the results of HLADQA10101-DQB10501, HLA-DQA10501-DQB10201, HLADQA10501-DQB10301 and HLA-DQA10102-DQB10602 are presented in Table I. For HLA-DR-based T cell epitopes, the results of HLA- DRB1*01:01, HLA- DRB3*01:01, HLA- DRB4*01:01, and HLA- DRB5*01:01 are also presented in Table I. As a result, Pla a 2 was predicted to have five T cell epitope sequences, 62–67, 86–91, 125–132, 217–222 and 343–350, as presented in Table II.