E. faecalis ATCC 29212 was purchased from American Type Culture Collection (Manassas, VA, USA) and grown on Brain heart infusion (BHI) agar (Becton Dickson, Sparks, MD, USA) at 37℃. For broth growth, bacterial cells were incubated in 10 mL of BHI broth for 12 hours at 37℃. Cell pellets were collected and kept at -80℃ until used. E. coli BL21 (DE3)/pLysS competent cells (Stratagene, La Jolla, CA, USA) were grown overnight on lysogeny broth (LB) agar plates or in LB broth containing kanamycin (30 µg/mL) at 37℃.

Genomic DNA from E. faecalis cells were extracted using a G-spin Genomic DNA Extraction Kit for Bacteria (iNtRON Biotechnology, Sungnam, Korea) according to manufacturer's instructions. Purified genomic DNA was then randomly sheared by sonication using a SONOPULS Ultrasonic homogenizer (Bandelin, Berlin, Germany) for 6 seconds at 100% power set of Sonic to generate DNA fragments sized of 1 - 5 kb. After fractionated by agarose gel electrophoresis to remove fragments that were smaller than 1 kb or bigger than 5 kb, DNA fragments of 1 - 5 kb were treated with End-It DNA End-Repair Kit (Epicentre, Madison, WI, USA) and subjected to phenol-chloroform extraction to eliminate any T4 polynucleokinase activity. Expression vector pET-30c (+) (Novagen, Madison, WI, USA) was digested with EcoRV, purified with QIAquick PCR Purification Kit (Qiagen, Valencia, CA, USA), and dephosphorylated with APex heat-labile alkaline phosphatase (Epicentre). A total of 250 ng of pET-30c (+)/EcoRV and blunt-ended E. faecalis DNA fragments at size of 1 - 5 kb were ligated at various molar ratios of vector to insert DNA (1:0.5, 1:1, 1:2, 1:5, and 1:10) using T4 DNA ligase (New England Biolab, Ipswich, MA, USA). The ligated mixture was used to transform E. coli DH5α (Takara, Daejon, Korea). Transformants were selected on LB plates containing kanamycin antibiotics. A total of 10 colonies were randomly selected and screened using colony PCR with T7 promoter primers. The frequency of self-ligation was lower than 20%. Purified recombinant DNA mixture was used to transform E. coli BL21 (DE3)/pLysS Competent Cells (Stratagene). Transformed cells were grown at 37℃ overnight on LB agar plates containing kanamycin (30 µg/mL).

Animal experimental protocols were reviewed and approved by the Institutional Animal Care and Use Committee, Chonnam National University (approval number: YB-2011-19). All surgical procedures were performed under sterile conditions. One 3 year old male beagle dog with weight of 10 kg was used for this study. Anesthetic induction was achieved by intravenous administration of thiopental (13.2 mg/kg body weight) followed by administration of 1 - 2% of isoflurane via an endotrachial tube. Additional local anesthesia of 2% lidocaine hydrochloride with 1:100,000 epinephrine was used to give a regional block. Preoperative radiograph was taken to determine any existing periodontal or endodontic lesion. After the induction, the second, third, and fourth premolars of each side of the upper jaw were used for inducing experimental endodontic infection as follows. A high-speed handpiece equipped with a round bur with saline irrigation was used to perforate a buccal surface of each tooth until it reached the pulp tissue. A total of 1.5 × 10⁵ E. faecalis cells were then injected into the pulp cavity using an insulin syringe to induce endodontic infection. The access opening on the buccal surface was subsequently covered with light-cured composite materials. After the surgical procedure, analgesic medication (3 mg/kg body weight, Ketopro, Unibiotech Co., Yesan, Korea) was administered for three days.

Three weeks after injecting bacterial cells into the pulp, a radiograph was taken again to assess the status of periapical lesion of infected teeth. At this time, the dog was anesthetized again using the same protocol described above. All infected premolars were extracted and hemisectioned. Infected pulp and root canal tissues were removed using a barbed broach. They were immediately dissolved in phosphate buffered saline (PBS). Bacterial cell pellets were obtained by centrifugation at ×4,000 g for 20 minutes as described previously.17 Purified cell pellets were immediately stored in -20℃. They were sent to a commercial vendor (Young In Frontier Co., Seoul, Korea) to produce hyperimmune antisera in rabbits. Based on optical density, it was determined that the bacterial cell pellets contained more than 5 × 10⁹ cells. The presence of E. faecalis was confirmed by visual inspection under a light microscope and sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis. The efficacy of produced antibodies was monitored by the enzyme-linked immunosorbent assay (ELISA) analysis.

To obtain a serum probe that reacted only to in vivo expressed E. faecalis antigens, hyperimmune rabbit sera were adsorbed with E. faecalis cells cultivated in vitro. Adsorptions were performed using whole bacterial cells. First, rabbit polyclonal antiserum was mixed with 1 × 10⁹ E. faecalis cells with slow agitation (30 rpm) at 4℃ for 1 hour. Cells were removed by centrifugation at ×5,000 g for 10 minutes and the serum (supernatant) was recovered. This step was repeated 4 times. After centrifugation, adsorbed serum was retrieved and stored at -80℃ until used. Cell lysates bound to nitrocellulose membrane were added into serum after removing whole cells and agitated overnight at 4℃. The serum was collected the next day. To remove antibodies that bound to E. coli proteins, the serum was agitated again with nitrocellulose membrane-bound E. coli BL21 harboring pET-30c(+) vector only.

To determine the adsorption efficiency, ELISA was performed as described previously.18 Briefly, a 96-well plate was coated with E. faecalis crude protein extracts at a concentration of 0.1 mg/100 mL in a coating buffer (0.1 M sodium carbonate, pH 9.6) overnight at 4℃. After washing, a blocking solution (2% skim milk in PBS, 300 µL) was added and incubated at 37℃ for 1 hour, followed by washing with PBS. The adsorbed rabbit antisera was added to each well using a serial dilution starting from 1:1,000 dilution and incubated at 37℃ for 2 hours. After the incubation, the plate was washed three times with 1X Tris buffered saline and Tween (TBS-Tween). Then, 50 mL of 1:10,000 diluted goat anti-human IgG conjugated with horse radish peroxidase (Abcam, Hanam, Korea) was added to each well and incubated at 37℃ for 1 hour. After washing the wells three times with TBS-Tween solution, TMB solution (GenDEPOT, Barker, TX, USA) was added to each well. After incubating at room temperature for 2 minutes, the reaction was stopped by adding 1 N H₂SO₄ (100 mL). Absorbance was measured at 450 nm using a 96-well plate reader (spectrometer).

E. coli recombinant clones were screened using the adsorbed hyperimmune rabbit sera as described previously.19 Briefly, kanamycin-resistant E. coli clones (approximately 200 per plate) were transferred to a nitrocellulose membrane. In vivo grown E. faecalis and E. coli containing pET-30c (+) were spotted onto the membrane as positive and negative control for screening, respectively. The membrane was then transferred (colony side up) to a LB agar plate containing 1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) to induce protein expression. After incubation at 37℃ for 3 hours, the membrane was removed and the attached cells were lysed by exposure to chloroform vapor. The membrane was then blocked with 5% skim milk in PBS for 1 hour, washed with PBS-Tween 20, and incubated overnight with the adsorbed rabbit hyperimmune serum (1:500) at 4℃. After the incubation, the membrane was washed and probed with goat anti-rabbit IgG conjugated with horse radish peroxidase (1:1,000) at room temperature for 1 hour. A 1-step chloronaphthol (4CN, Thermo Scientifc, Waltham, MA, USA) was used as substrate to identify reactive clones. Positive clones showing reactivity were isolated from the original (master) plate. They were grown on LB-kanamycin plates. Their reactivity was confirmed again using a 4CN as substrate.

Plasmid DNAs from positive clones were purified using a QIAprep Spin Miniprep Kit (Qiagen, Valencia, CA, USA). DNA sequences of inserted DNA fragments from E. faecalis were determined by direct sequencing using a T7 promoter primer with an ABI Prism 377 automatic DNA sequencer by double-strand dideoxy chain termination method at GenoTech Corp. (Daejeon, Korea). Identified gene sequences were compared to DNA and protein databases using a BLAST program (http://www.ncbi.nlm.nih.gov) and analyzed by Vector NTI Software (Invitrogen, Carlsbad, CA, USA). Functional classification of identified antigens was based on published studies of identified proteins of E. faecalis if available.

SDS-PAGE analysis confirmed the presence of bacterial cells in the sample acquired from experimental endodontic infection (Figure 1). The hyperimmune rabbit antisera produced against E. faecalis infected pulp and periapical tissues were successively adsorbed against E. faecalis cells grown in vitro. The adsorption efficiency was determined by examining immuno-reactivity of antisera using ELISA after each adsorption step with in vitro grown E. faecalis. Fully adsorbed serum exhibited a significant lower reactivity to in vitro grown E. faecalis (Figure 2), suggesting that the adsorption process efficiently removed antibodies against most E. faecalis proteins produced during in vitro cultivation.

Plasmid DNAs of ten colonies randomly selected from the library were subjected to direct DNA sequencing to confirm the presence of the DNA inserts present in the recombinant plasmids. Of the 10 colonies, eight had DNA inserts in the pET-30c (+) vector. A total of approximately 4,500 recombinant clones were screened using the adsorbed pooled serum. In vivo grown E. faecalis cells and E. coli cells containing only pET-30c (+) placed on the nitrocellulose membrane was used as positive and negative control, respectively, for the immunoscreening. Initial screening resulted in the detection of 30 immuno-reactive clones. Of the 30 clones, nineteen reproducibly exhibited reactivity with the adsorbed serum. Therefore, they were subjected to further DNA sequencing.

Sequencing of the DNA inserts of the 19 positive clones revealed 16 open reading frames (ORFs). Sequences of the remaining three clones could not be obtained. Sequences of the 16 clones were analyzed using BLAST program. CMAT-identified in vivo-induced genes and their predicted functions are listed in Table 1. Many determined gene sequences contained portions of ORFs for putative genes. The size of ORF varied from 663 bp to 2,911 bp. In addition, multiple ORFs were not identified. Sequence analysis revealed that many of these ORFs were enzymes implicated in housekeeping functions with intermediary metabolism. In addition, conserved hypothetical proteins with unknown function and putative membrane protein were also identified, along with copper resistance proteins.