Abstract Flagellin is a classical pathogen-associated molecular pattern that can evoke a robust immune response. We have demonstrated previously that three full-length flagellins of Treponema pallidum, namely FlaB1, FlaB2 and FlaB3, did have diagnostic value in the serodiagnosis of syphilis. Here, we selected and constructed three recombinant fragments of each complete FlaB, both the conserved N-terminal and the C-terminal region, and the middle variable part, with the goal of exploring fragments unique to Treponema pallidum for use as antigen targets in a fragment-based serological test. The diagnostic performance of fragments was evaluated using different panels of serum specimens (= 332) by indirect IgG enzyme-linked immunosorbent assay. The data showed that all the conserved fragments exhibited excellent sensitivities (91.1–95.0%) but poor specificities (64.1–78.4%), while the three middle regions demonstrated higher sensitivities and specificities for detecting IgG antibody, with 92.7% and 96.1% for FlaB1M (‘B1M’), 91.6% and 94.8% for B2M, and 95.0% and 100% for B3M, respectively. In comparison, the sensitivity and specificity of Architect Syphilis TP was found to be 95.5% and 94.8%, respectively. These findings revealed that the middle portion of each FlaB had epitopes specific for Treponema pallidum and identified B3M as a promising candidate antigen for the serodiagnosis of syphilis. Treponema pallidum, FlaB protein fragments, serodiagnosis, syphilis INTRODUCTION Syphilis is caused by infection with Treponema pallidum subsp. pallidum, a not-yet-cultivable spiral-shaped bacterium that is usually transmitted by sexual contact with an infected partner or by an infected pregnant woman to her fetus (Stamm and Drapp 2014). Despite the efficiency of treatment with antibiotics, syphilis remains a global health concern, with an estimated annual incidence of 11 million new cases worldwide among those aged 15–49 years (World Health Organization 2011). Several regions, such as North America, Western Europe, China and Australia, have experienced the re-emerging of syphilis (Stamm 2016). In recent decades, the incidence of syphilis among men who have sex with men has been on the rise (Woolston, Dhanireddy and Marrazzo 2016; Ong et al.2017). In addition, the lesions of homosexual men with syphilis may increase the risk for acquiring and transmitting the human immunodeficiency virus (HIV) (Buchacz et al.2008). The inability to cultivate Treponema pallidum makes the diagnosis of syphilis challenging and has forced laboratories to seek alternative approaches for diagnostic application. The direct tests, dark-field microscopy or polymerase chain reaction (PCR), are useful aids in the diagnosis of early primary syphilis. Both tests, however, are highly operator-dependent and inapplicable for clinics lacking the fairly expensive instruments. Thus, the diagnosis of syphilis is based mainly on serology. As no single serologic test is sufficient, serodiagnosis of syphilis requires the detection of non-treponemal antibodies and treponemal antibodies. The non-treponemal antibodies, primarily directed against phospholipids, can be measured by the rapid plasma reagin (RPR) test, the Venereal Disease Research Laboratory (VDRL) test, and the toluidine red unheated serum test (TRUST). The treponemal antibodies, mainly directed against T. pallidum polypeptides, can be detected by the fluorescent treponemal antibody absorbed (FTA-ABS) test, T. pallidum hemagglutination assay (TPHA) or the T. pallidum particle agglutination (TPPA) assay. Traditionally, serum samples are first screened with a non-treponemal test and then reactive samples are confirmed with a treponemal test (Tsang et al.2007; Hoover and Radolf 2011; Morshed and Singh 2015). While this algorithm is cost-effective and reliably correlates the results with infection status (Samoff et al.2009; Binnicker 2012), the non-treponemal test as a screening assay has several limitations, including lower specificity, manual operation and subjective interpretation of test results (Binnicker 2012). With the advent of automated treponemal tests, such as enzyme immunoassays (EIAs) and chemiluminescence immunoassays, utilizing individual or a combination of recombinant lipoproteins TpN15 (Tp0171), TpN17 (Tp0435), TpN44 (Tp0768, TmpA) and TpN47 (Tp0574), some laboratories have begun to employ a reverse algorithm (Hoover and Radolf 2011; Smith et al.2013). In this approach, specimens are first screened with treponemal tests (e.g. EIAs), positive samples are then tested with a quantitative non-treponemal assay (e.g. RPR), and then discordant samples (e.g. EIA reactive but RPR non-reactive) are re-screened with a second and different treponemal assay (Centers for Disease Control and Prevention 2011; Lipinsky et al.2012; Loeffelholz and Binnicker 2012). However, this algorithm has shown higher false-reactive rates than the traditional one in populations with a low prevalence of syphilis (Centers for Disease Control and Prevention 2011; Binnicker, Jespersen and Rollins 2012). Thus, it is of paramount importance to evaluate highly sensitive and specific recombinant antigens for the serodiagnosis of syphilis. Flagellin, a classic pathogen-associated molecular pattern (PAMP) protein, has the capacity to trigger robust innate and adaptive immune responses, thus being considered a potent adjuvant for vaccines and immunotherapy (Mizel and Bates 2010; Nguyen et al.2011, 2013; Zgair and Chhibber 2012). Flagellin B (FlaB), including FlaB1 (Tp0868), FlaB2 (Tp0792) and FlaB3 (Tp0870), are inner core proteins of the perplasmic Treponema pallidum flagella (Liu et al.2010). Our previous study showed that each full-length FlaB does show promise as a valuable candidate antigen in the serodiagnosis of syphilis (Jiang et al.2016). However, the specificity may be insufficient due to the sequence homology between the FlaB protein of T. pallidum and a broad class of flagellins present in a wide variety of bacteria (Norris et al.1988). Studies have demonstrated that sequence homology is mainly located at the N- and C-terminal regions of these flagellins (Joys 1985; Norris et al.1988; Gassmann et al.1991; Norris 1993). Thus, with the aim of seeking highly sensitive and extremely specific antigens, we selected and designed three fragments of each FlaB by using bioinformatics software. The diagnostic potential of the recombinant fragments were evaluated by screening sera from various stages of syphilis, potentially cross-reactive diseases and healthy individuals. For direct comparison, the performance of Architect Syphilis TP was also tested against the same serum panels. MATERIALS AND METHODS Cloning, expression and purification of FlaB protein fragments The recombinant plasmids of the three FlaB genes used in this study were prepared in our previous study (Jiang et al.2016). Based on homology analysis, each full-length FlaB protein was fragmented into three continuous parts, i.e. the conserved N-terminal and C-terminal regions, and the middle variable portion. These nine fragments were cloned and expressed as follows. The regions encoding each fragment were amplified by PCR using the corresponding recombinant plasmids of the complete FlaB as a template. The amplification profile was 94°C for 5 min, 30 cycles consisting of 94°C for 30 s, 67°C for 30 s and 72°C for 2 min followed by a 72°C final extension for 5 min. The resulting amplicons were cloned into the prokaryotic expression vector pET-28a (Novagen, Merck Millipore, Darmstadt, Germany), transformed initially into Escherichia coli strain JM109 (Invitrogen, Carlsbad, CA) and selected on Luria-Bertani agar plates containing 50 μg ml−1 kanamycin. The positive clones were confirmed by PCR, restriction enzyme identification, and DNA sequencing. The respective specific primers, restriction endonucleases, amplicon size and predicted molecular weight are shown in Table 1. Table 1. Primers and restriction endonucleases used for cloning of FlaB fragments. FlaB fragment Primer sequences (5΄–3΄) a Restriction endonucleases Amplicon size (bp) Corresponding amino acids Predicted molecular weight (kDa) B1N F:CGCGGATCCATGATTATCAATCACAAC R:CCGCTCGAGTTAGGAGAAGCGGCCCGTGAG BamHI XhoI 426 1–141 15.4 B1M F:CGCGGATCCATGCGCACTGAAGGTGAGAACG R:CCGCTCGAGTTAGCTCTTGTTGGCCGAGTCT BamHI XhoI 180 142–199 6.28 B1C F:CGCGGATCCATGATCGGCACCATCGATGCTG R:CCGCTCGAGTTACCGGAGAATTGAGAG BamHI XhoI 267 200–286 9.5 B2N F:CGCGGATCCATGATCATCAATCACAACATG R:CCCAAGCTTCTACGCGAAGCGACCAGTGAG BamHI HindIII 426 1–141 15.4 B2M F:CGCGGATCCATGCGTCAAGGCGGGGAGAAC R:CCCAAGCTTCTATGCGCGGTTGGCCTTTTC BamHI HindIII 180 142–199 6.28 B2C F:CGCGGATCCATGATCGGTACGCTTGATCAG R:CCGCTCGAGCTAACGCAAGAGGCTTAGAAC BamHI XhoI 267 200–286 9.5 B3N F:CGCGGATCCATGATTATCAATCACAACATG R:CCGGAATTCTTAGGAGAAGCGGCCCGTGAG BamHI EcoRI 426 1–141 15.4 B3M F:CGCGGATCCATGCGCGAGTCTGCCCTTGGG R:CCGGAATTCTTAGACCTTGTTCGCCCCGTC BamHI EcoRI 180 142–199 6.28 B3C F:CGCGGATCCATGATCGGTACGCTTGATAGC R:CCGGAATTCTTACTGCATCAAGCGGAG BamHI EcoRI 267 200–285 9.5 FlaB fragment Primer sequences (5΄–3΄) a Restriction endonucleases Amplicon size (bp) Corresponding amino acids Predicted molecular weight (kDa) B1N F:CGCGGATCCATGATTATCAATCACAAC R:CCGCTCGAGTTAGGAGAAGCGGCCCGTGAG BamHI XhoI 426 1–141 15.4 B1M F:CGCGGATCCATGCGCACTGAAGGTGAGAACG R:CCGCTCGAGTTAGCTCTTGTTGGCCGAGTCT BamHI XhoI 180 142–199 6.28 B1C F:CGCGGATCCATGATCGGCACCATCGATGCTG R:CCGCTCGAGTTACCGGAGAATTGAGAG BamHI XhoI 267 200–286 9.5 B2N F:CGCGGATCCATGATCATCAATCACAACATG R:CCCAAGCTTCTACGCGAAGCGACCAGTGAG BamHI HindIII 426 1–141 15.4 B2M F:CGCGGATCCATGCGTCAAGGCGGGGAGAAC R:CCCAAGCTTCTATGCGCGGTTGGCCTTTTC BamHI HindIII 180 142–199 6.28 B2C F:CGCGGATCCATGATCGGTACGCTTGATCAG R:CCGCTCGAGCTAACGCAAGAGGCTTAGAAC BamHI XhoI 267 200–286 9.5 B3N F:CGCGGATCCATGATTATCAATCACAACATG R:CCGGAATTCTTAGGAGAAGCGGCCCGTGAG BamHI EcoRI 426 1–141 15.4 B3M F:CGCGGATCCATGCGCGAGTCTGCCCTTGGG R:CCGGAATTCTTAGACCTTGTTCGCCCCGTC BamHI EcoRI 180 142–199 6.28 B3C F:CGCGGATCCATGATCGGTACGCTTGATAGC R:CCGGAATTCTTACTGCATCAAGCGGAG BamHI EcoRI 267 200–285 9.5 aF:forward primer; R: reverse primer; restriction enzyme recognition sequences are indicated by underlining. View Large In order to express the nine FlaB fragments, the respective recombinant plasmids were re-transformed into E.coli BL21 (DE3) (Novagen). Log phase cultures of the selected clones were induced for 6 h at 37°C with 0.5 mM isopropyl-β-d-thiogalactopyranoside. Bacteria were harvested by centrifugation, lysed in a buffer containing 50 mM Tris-HCl (pH 7.8), 250 mM NaCl, 10 mM imidazole, 20% glycerol, and 1% Triton X-100, and sonicated on ice. Both the supernatant and the pellet were analysed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). These recombinant His-tagged fragments were expressed as inclusion bodies and purified by affinity chromatography using Ni-nitrilotriacetic acid (NTA) beads (QIAGEN, Hilden, Germany). Briefly, the pellets of broken induced bacteria were firstly dissolved in buffer A (100 mM NaH2PO4, 10 mM Tris-HCl, 8 M urea, pH 8.0) for 6 h on ice and then added into an NTA agarose column shaking on ice for 5 h. After being washed by buffer A with different pH (8.0, 6.3, 5.9 or 4.5) in sequence, and eluted with buffer A (pH 4.5), these fragments were subsequently dialyzed using renaturation solution (50 mM Tris, 50 mM NaCl, 1 mM glutathione (GSH), 0.5 mM EDTA, 0.2 mM DTT, 0.1 mM oxidized glutathione (GSSG), 30% glycerol, pH 8.35) and then concentrated by PEG 6000 (Solarbio, Beijing, China), the concentrations of recombinant fragments were estimated using bicinchoninic acid (BCA) protein assay kit (Pierce, Rockford, IL). The purity and molecular size were determined by means of SDS-PAGE on 15% acrylamide gels. Western blotting analysis of nine fragments was done with anti-His monoclonal antibody (1:1000 dilution) and rabbit anti-FlaBFL (1:2000; raised against the corresponding full-length FlaB) antibodies as standard protocols. Serum panels This study was approved by the Human Ethics Committee of the University of South China and written informed consent was obtained from all the participants. Serum samples collected are divided into three panels (A–C), as follows: panel A, healthy control, included 30 samples submitted to The Second Affiliated Hospital, University of South China (Hengyang, China) for routine screening for syphilis. Then, 179 syphilitic sera belonging to panel B were obtained from patients with primary (= 44), secondary (= 70) or latent (= 65) infections attending The First Affiliated Hospital, University of South China (Hengyang, China) and The First People's Hospital of Huaihua (Huaihua, China) during September 2016 and December 2016. The staging of syphilis was done according to the criteria proposed in the literature (Jiang et al.2016; Xu et al.2016): primary syphilis, typical hard chancre and positive serologic test results; secondary syphilis, generalized rash and lymphadenopathy, and positive serologic test results; latent syphilis, no clinical manifestations and no history of treatment during the previous 2 years, and positive serologic test results. Panel C, designated as disease control, consisted of 123 serum samples collected from patients suffering from some infections probably resulting in cross-reactivity in syphilis serodiagnostic assays, including Lyme disease (= 30), leptospirosis (= 30) and Helicobacter pylori infection (= 8), hepatitis B virus infection (= 26), hepatitis C virus infection (= 11), cytomegalovirus infection (= 12) and Epstein–Barr virus infection (= 6). In particular, the causative agents of the former three conditions are flagellated bacteria, of which the former two, Borrelia burgdorferi and Leptospira spp., are also closely related spirochetes. Sera from patients diagnosed with Lyme disease were obtained from the Chinese Centre for Disease Control and Prevention (Beijing, China). Leptospirosis sera were obtained from the Hunan Provincial Centre for Disease Control and Prevention (Changsha, China), and other disease control samples were obtained from The First Affiliated Hospital, University of South China (Hengyang, China). The diagnostic tests used to confirm each of the control samples are listed in Table 2. Table 2. Serological tests used to confirm the serum samples. Microorganism Serological testa Manufacturer Treponema pallidum Rapid plasma reagin test KeHua Bio-Engineering Inc., Shanghai, China Treponema pallidum T. pallidum particle agglutination assay Fujirebio, Tokyo, Japan Borrelia burgdorferi IgG western blotting MarDx Diagnostics, Inc., Carlsbad, CA Leptospira spp. Leptospira IgM assay DRG, Marburg, Germany Helicobacter pylori H. pylori IgG assay DRG, Marburg, Germany Hepatitis B virus Total anti-HBs or anti-HBc test LivZon Diagnostics, Inc., Zhuhai, China Hepatitis C virus (HCV) Anti-HCV IgG assay LivZon Diagnostics, Inc., Zhuhai, China Cytomegalovirus Cytomegalovirus IgG assay DRG, Marburg, Germany Epstein–Barr virus (EBV) Anti-EBV IgG/IgM assay DRG, Marburg, Germany Microorganism Serological testa Manufacturer Treponema pallidum Rapid plasma reagin test KeHua Bio-Engineering Inc., Shanghai, China Treponema pallidum T. pallidum particle agglutination assay Fujirebio, Tokyo, Japan Borrelia burgdorferi IgG western blotting MarDx Diagnostics, Inc., Carlsbad, CA Leptospira spp. Leptospira IgM assay DRG, Marburg, Germany Helicobacter pylori H. pylori IgG assay DRG, Marburg, Germany Hepatitis B virus Total anti-HBs or anti-HBc test LivZon Diagnostics, Inc., Zhuhai, China Hepatitis C virus (HCV) Anti-HCV IgG assay LivZon Diagnostics, Inc., Zhuhai, China Cytomegalovirus Cytomegalovirus IgG assay DRG, Marburg, Germany Epstein–Barr virus (EBV) Anti-EBV IgG/IgM assay DRG, Marburg, Germany aHBs, hepatitis B surface antigen; HBc, hepatitis B core antigen. View Large All serum panels were screened by RPR test and TPPA assay in a blinded manner according to the manufacturers’ instructions. The serum samples were stored in aliquots at –80°C prior to use and underwent only three freeze–thaw cycles. Fragment-based indirect IgG enzyme-linked immunosorbent assay The indirect IgG enzyme-linked immunosorbent assays (ELISAs) were performed as previously described (Jiang et al.2016). The optimal concentrations of FlaB fragments and the ideal dilutions of sera were determined by a chessboard titration approach. Ninety-six-well plates (Costar, Corning, NY) were coated overnight at 4°C with recombinant fragments at corresponding optimal concentrations in coating buffer (0.05 M carbonate buffer, pH 9.6). After washing three times with PBST (phosphate-buffered saline with 0.05% Tween-20), plates were then blocked with PBSTM (PBST with 5% skim milk) for 2 h at 37°C. Serum samples were diluted at 1:200 with PBSTM and added to each well. They were then incubated for 1 h at 37°C. Each plate included samples from patients and healthy individuals, and all samples were tested in duplicate. After being washed five times, they were incubated for 1 h at 37°C with 1:10 000 dilution of horseradish peroxidase-conjugated goat anti-human IgG (GeneTex, San Antonio, TX). Finally, a chromogenic peroxidase substrate, tetramethylbenzidine (TMB) (Sigma, St. Louis, MO), was added to each well and the reaction stopped after incubation for 15 min at 37°C in darkness by adding 2 M sulfuric acid. The absorbance at 450 nm was measured using an ELISA microplate reader (Bio-Rad, Hercules, CA). Architect Syphilis TP All serum panels were tested in duplicate on an automatic Architect i2000SR analyzer (Abbott, Chicago, IL) according to the manufacturer's instructions. Architect Syphilis TP is a two-step chemiluminescent immunoassay that employs chemiluminescent microparticle immunoassay technology for qualitative detection of anti-T. pallidum antibodies in human serum or plasma (Marangoni et al.2013). The results are measured as relative light units (RLUs), the level of which is associated with the amount of total IgG and/or IgM in the tested sera. Samples with an RLU/cutoff value of ≥1.00 were determined as positive, while those with an RLU/cutoff value of <1.00 were considered as negative. Statistics GraphPad Prism 5.0 software (GraphPad, San Diego, CA) was used for statistical analysis of the data. The cutoff values were presented as the mean plus two times the standard deviation of the absorbance values for the uninfected individuals (mean + 2 SD). Samples with absorbance values less than or equal to the cutoff value were defined as negative. A chi-square test was performed to directly compare positive and negative results for each fragment. Differences were considered statistically significant at two-tailed P values of <0.05. The percentage agreement, kappa (κ) coefficients, and sensitivity and specificity with 95% confidence intervals (CIs) were calculated. Result agreement by kappa values is classified as almost perfect (0.81–1.0), substantial (0.61–0.8), moderate (0.41–0.6), fair (0.21–0.4), slight (0–0.2), or poor (<0) (Landis and Koch 1977). RESULTS Preparation and identification of recombination FlaB fragments The amino acid sequences of three full-length FlaB proteins of T. pallidum have 33–40% identity to that of B. burgdorferi, Leptospira spp., H. pylori, E. coli and S. typhimurium. However, the homology of the N- and C-terminal amino acid sequences of flagellins from the above bacteria exceeded 50% and 60%, respectively. Thus, the homology is greatest at the amino and carboxyl termini of each FlaB, with the middle region showing the most variability. The gene fragments chosen based on the homology analysis results were successfully amplified by PCR utilizing specific primers (Fig. 1). All His-tagged recombinant fragments were expressed as insoluble inclusion bodies and then subjected to immobilized metal affinity chromatography using Ni-NTA beads. Each purified fragment appeared as a single band and was judged to have a purity of >95% via SDS-PAGE analysis (Fig. 2), and western blotting showed that each fragment reacted positively to anti-His monoclonal antibody and corresponding anti-FlaBFL antibodies (Fig. 3). Figure 1. View largeDownload slide PCR amplification of the nine selected flaB gene fragments. Lane1, 100 bp DNA ladder; lanes 2–10, every three fragments corresponding to flaB1, flaB2 and flaB3, respectively. Figure 1. View largeDownload slide PCR amplification of the nine selected flaB gene fragments. Lane1, 100 bp DNA ladder; lanes 2–10, every three fragments corresponding to flaB1, flaB2 and flaB3, respectively. Figure 2. View largeDownload slide SDS-PAGE analysis of purified recombinant FlaB protein fragments. Lane 1, pre-stained protein marker; lanes 2–10, every three fragments corresponding to FlaB1, FlaB2 and FlaB3, respectively. Figure 2. View largeDownload slide SDS-PAGE analysis of purified recombinant FlaB protein fragments. Lane 1, pre-stained protein marker; lanes 2–10, every three fragments corresponding to FlaB1, FlaB2 and FlaB3, respectively. Figure 3. View largeDownload slide Western blot identification of nine FlaB protein fragments. The upper panel shows the western blot analysis of target FlaB fragments with corresponding rabbit anti-FlaBFL sera (FlaBFL, namely the full-length FlaB). The lower panel reveals the western blot analysis of the same fragments with anti-His monoclonal antibody. Figure 3. View largeDownload slide Western blot identification of nine FlaB protein fragments. The upper panel shows the western blot analysis of target FlaB fragments with corresponding rabbit anti-FlaBFL sera (FlaBFL, namely the full-length FlaB). The lower panel reveals the western blot analysis of the same fragments with anti-His monoclonal antibody. Immunoreactivity of the recombinant FlaB fragments The sensitivity of each recombinant FlaB fragment was evaluated by screening sera from healthy individuals (panel A) and from different stages of syphilis (panel B). The cutoff values calculated were significantly higher for terminal fragments compared to that of middle regions, of which the minimum value for B3M was 0.27 (Table 4). The overall sensitivity for all fragments ranged from 91.1 to 95.0% (Table 4). No significant differences were detected among the nine fragments (P > 0.05). To determine the specificities of the fragment-based ELISA, serum samples collected from healthy individuals (panel A) and those with potentially cross-reactive disease (panel C, or disease control) were tested. It is noteworthy that several serum samples from patients with cytomegalovirus and Epstein–Barr virus infection gave faint cross-reactivity in terminal fragment-based ELISA, whereas some samples belonging to other diseases yielded slightly strong or mild positive responses. These terminal fragments could react with at least five kinds of disease control serum samples; in particular, B2C, the C-terminal of FlaB2, was almost reactive with samples of all kinds, exhibiting the poorest specificity (Table 3). Depending on the six purified recombinant terminal fragments used, 1–3 of 12 patients with cytomegalovirus infection, 3–7 of 11 patients with hepatitis C virus infection, 0–2 of 6 patients with Epstein–Barr virus infection, 5–15 of 26 patients with hepatitis B virus infection, 1–3 of 8 patients with H. pylori infection, 9–18 of 30 patients with Lyme disease, 3–6 of 30 patients with leptospirosis, and 1–4 of 30 uninfected individuals were positive (Table 3). The overall specificity for these six fragments ranged from 64.1 to 78.4% (Table 4). With regard to the three variable middle fragments, all of which seldom yielded cross-reactivity with the negative control serum samples (panels A and C), the overall specificity was 96.1% for B1M, 94.8% for B2M and 100% for B3M. There are significant differences between B3M and the other fragments (P < 0.05). Table 3. Serum IgG antibody reactivity with recombinant proteins. Seraa No. tested No. (%) of sera reactive with recombinant protein Architect Syphilis B1N B1M B1C B2N B2M B2C B3N B3M B3C TP PS 44 43 (97.7) 43 (97.7) 38 (86.4) 44 (100.0) 39 (88.6) 37 (84.1) 42 (95.5) 41 (93.2) 39 (88.6) 41 (93.2) SS 70 67 (95.7) 67 (95.7) 65 (92.9) 63 (90.0) 66 (94.3) 70 (100.0) 64 (91.4) 65 (92.9) 62 (88.6) 68 (97.1) LS 65 59 (90.8) 56 (86.2) 62 (95.4) 63 (96.9) 59 (90.8) 61 (93.8) 63 (96.9) 64 (98.5) 62 (95.4) 62 (95.4) CMV 12 2 (16.7) 0 1 (8.3) 3 (25.0) 0 3 (25.0) 3 (25.0) 0 2 (16.7) 0 HCV 11 3 (27.3) 2 (18.2) 7 (63.6) 5 (45.5) 3 (27.3) 6 (54.5) 3 (27.3) 0 5 (45.5) 0 EBV 6 2 (33.3) 0 0 2 (33.3) 0 1 (16.7) 0 0 0 1 (16.7) HBV 26 5 (19.2) 3 (11.5) 12 (46.2) 7 (26.9) 4 (15.4) 15 (57.7) 10 (38.5) 0 13 (50.0) 2 (7.7) Hp 8 2 (25.0) 0 3 (37.5) 1 (12.5) 0 2 (25.0) 2 (25.0) 0 2 (25.0) 0 LD 30 12 (40.0) 1 (3.3) 16 (53.3) 14 (46.7) 0 18 (60.0) 15 (50.0) 0 9 (30.0) 2 (6.7) Le 30 5 (16.7) 0 4 (13.3) 3 (10.0) 1 (3.3) 6 (20.0) 5 (16.7) 0 5 (16.7) 0 U 30 2 (6.7) 0 1 (3.3) 3 (10.0) 0 4 (13.3) 3 (10.0) 0 3 (10.0) 3 (10.0) Seraa No. tested No. (%) of sera reactive with recombinant protein Architect Syphilis B1N B1M B1C B2N B2M B2C B3N B3M B3C TP PS 44 43 (97.7) 43 (97.7) 38 (86.4) 44 (100.0) 39 (88.6) 37 (84.1) 42 (95.5) 41 (93.2) 39 (88.6) 41 (93.2) SS 70 67 (95.7) 67 (95.7) 65 (92.9) 63 (90.0) 66 (94.3) 70 (100.0) 64 (91.4) 65 (92.9) 62 (88.6) 68 (97.1) LS 65 59 (90.8) 56 (86.2) 62 (95.4) 63 (96.9) 59 (90.8) 61 (93.8) 63 (96.9) 64 (98.5) 62 (95.4) 62 (95.4) CMV 12 2 (16.7) 0 1 (8.3) 3 (25.0) 0 3 (25.0) 3 (25.0) 0 2 (16.7) 0 HCV 11 3 (27.3) 2 (18.2) 7 (63.6) 5 (45.5) 3 (27.3) 6 (54.5) 3 (27.3) 0 5 (45.5) 0 EBV 6 2 (33.3) 0 0 2 (33.3) 0 1 (16.7) 0 0 0 1 (16.7) HBV 26 5 (19.2) 3 (11.5) 12 (46.2) 7 (26.9) 4 (15.4) 15 (57.7) 10 (38.5) 0 13 (50.0) 2 (7.7) Hp 8 2 (25.0) 0 3 (37.5) 1 (12.5) 0 2 (25.0) 2 (25.0) 0 2 (25.0) 0 LD 30 12 (40.0) 1 (3.3) 16 (53.3) 14 (46.7) 0 18 (60.0) 15 (50.0) 0 9 (30.0) 2 (6.7) Le 30 5 (16.7) 0 4 (13.3) 3 (10.0) 1 (3.3) 6 (20.0) 5 (16.7) 0 5 (16.7) 0 U 30 2 (6.7) 0 1 (3.3) 3 (10.0) 0 4 (13.3) 3 (10.0) 0 3 (10.0) 3 (10.0) aPS, primary syphilis; SS, secondary syphilis; LS, latent syphilis; CMV, cytomegalovirus infection; HCV, hepatitis C virus infection; EBV, Epstein–Barr virus infection; HBV, hepatitis B virus infection; Hp, Helicobacter pylori infection; LD, Lyme disease; Le, leptospirosis; U, uninfected individuals. View Large Table 4. Cutoff values, sensitivities and specificities of proteins in the detection of syphilis.a Cutoff Sensitivity, % Specificity, % Protein value (95% CI) (95% CI) B1N 0.5556 94.4 (90.0–96.9) 78.4 (71.3–84.2) B1M 0.3645 92.7 (88.0–95.7) 96.1 (91.7–98.2) B1C 0.4831 92.2 (87.3–95.3) 71.2 (63.6–77.8) B2N 0.4471 95.0 (90.7–97.3) 75.2 (67.8–81.3) B2M 0.4346 91.6 (86.6–94.9) 94.8 (90.0–97.3) B2C 0.5107 93.9 (89.3–96.5) 64.1 (56.2–71.2) B3N 0.4412 94.4 (90.0–96.9) 73.2 (65.7–79.6) B3M 0.2700 95.0 (90.7–97.3) 100.0 (97.6–100.0) B3C 0.5495 91.1 (86.0–94.4) 74.5 (67.1–80.8) Architect Syphilis TP – 95.5 (91.4–97.7) 94.8 (90.0–97.3) Cutoff Sensitivity, % Specificity, % Protein value (95% CI) (95% CI) B1N 0.5556 94.4 (90.0–96.9) 78.4 (71.3–84.2) B1M 0.3645 92.7 (88.0–95.7) 96.1 (91.7–98.2) B1C 0.4831 92.2 (87.3–95.3) 71.2 (63.6–77.8) B2N 0.4471 95.0 (90.7–97.3) 75.2 (67.8–81.3) B2M 0.4346 91.6 (86.6–94.9) 94.8 (90.0–97.3) B2C 0.5107 93.9 (89.3–96.5) 64.1 (56.2–71.2) B3N 0.4412 94.4 (90.0–96.9) 73.2 (65.7–79.6) B3M 0.2700 95.0 (90.7–97.3) 100.0 (97.6–100.0) B3C 0.5495 91.1 (86.0–94.4) 74.5 (67.1–80.8) Architect Syphilis TP – 95.5 (91.4–97.7) 94.8 (90.0–97.3) aFor each assay, the cutoff value, sensitivity and specificity are calculated. The 95% confidence interval (CI) of sensitivity and specificity are recorded in the parentheses. View Large Diagnostic performance of Architect Syphilis TP assay For comparison, the diagnostic performance of Architect Syphilis TP chemiluminescent immunoassay (CIA) was assayed against the same sample set. This automatic CIA exhibited an overall sensitivity of 95.5%, with sensitivities of 93.2%, 97.1% and 95.4% for the detection of primary, secondary and latent infection, respectively. The overall specificity was 94.8%, with three serum samples from uninfected control, two samples from Lyme disease, one sample from a patient infected with Epstein–Barr virus, and two samples from patients infected with hepatitis B virus scoring positive. Direct comparison of recombinant fragment-based IgG ELISA and Architect Syphilis TP to TPPA assay Following the testing of 332 serum samples, the percentage of agreement and κ values of the fragment-based IgG ELISA and the Architect Syphilis TP were compared with the TPPA assay, and the results are shown in Table 5. For the six terminal fragments, the overall percentage of agreements and the κ values ranged from 80.1 to 87.0% and 0.59 to 0.74, respectively. In contrast, the three middle fragments exhibited high consistency with the TPPA assay, with the overall percentage of agreements and the κ values ranging from 93.1 to 97.3% and 0.86 to 0.95, respectively. Finally, the Architect Syphilis TP exhibited 95.2% agreement and a κ value of 0.90. These results indicated that both the middle fragment-based ELISA and the Architect Syphilis TP had almost perfect agreement. Table 5. Comparison of recombinant FlaB fragment-based ELISA and Architect Syphilis TP to TPPAa assay. Assay TPPA Agreement, % (95% CI) κ value Positive Negative B1N Positive 169 33 87.0 (84.0–90.1) 0.74 Negative 10 120 B1M Positive 166 6 94.3 (92.4–96.1) 0.89 Negative 13 147 B1C Positive 165 44 82.5 (79.0–86.1) 0.64 Negative 14 109 B2N Positive 170 38 85.8 (82.6–89.0) 0.71 Negative 9 115 B2M Positive 164 8 93.1 (91.0–95.1) 0.86 Negative 15 145 B2C Positive 168 55 80.1 (76.2–84.1) 0.59 Negative 11 98 B3N Positive 169 41 84.6 (81.3–88.0) 0.69 Negative 10 112 B3M Positive 170 0 97.3 (96.0–98.6) 0.95 Negative 9 153 B3C Positive 163 39 83.4 (80.0–86.8) 0.66 Negative 16 114 Architect Syphilis TP Positive 171 8 95.2 (93.4–96.9) 0.90 Negative 8 145 Assay TPPA Agreement, % (95% CI) κ value Positive Negative B1N Positive 169 33 87.0 (84.0–90.1) 0.74 Negative 10 120 B1M Positive 166 6 94.3 (92.4–96.1) 0.89 Negative 13 147 B1C Positive 165 44 82.5 (79.0–86.1) 0.64 Negative 14 109 B2N Positive 170 38 85.8 (82.6–89.0) 0.71 Negative 9 115 B2M Positive 164 8 93.1 (91.0–95.1) 0.86 Negative 15 145 B2C Positive 168 55 80.1 (76.2–84.1) 0.59 Negative 11 98 B3N Positive 169 41 84.6 (81.3–88.0) 0.69 Negative 10 112 B3M Positive 170 0 97.3 (96.0–98.6) 0.95 Negative 9 153 B3C Positive 163 39 83.4 (80.0–86.8) 0.66 Negative 16 114 Architect Syphilis TP Positive 171 8 95.2 (93.4–96.9) 0.90 Negative 8 145 aTreponema pallidum particle agglutination assay. View Large DISCUSSION An ideal serological test is vital to syphilis diagnosis due to the fact that T. pallidum cannot be cultured in vitro (Mattei et al.2012; Henao-Martinez and Johnson 2014). Firstly, the test must be widely available, cheap and easy to perform. Secondly, they should meet the needs of high specificity and sensitivity simultaneously. High sensitivity is the pivotal element required for any diagnostic tests, but specificity is also of considerable significance, since false-positive results may cause various unpleasant situations for those involved, such as unnecessary examinations and treatment. Currently, many serological kits for syphilis serodiagnosis are suffering from insufficient specificity that arises from the use of whole-cell extracts or whole proteins. The former contain a variety of proteins and other macromolecules, most of which may influence the results of the assay. The latter contain epitopes that may yield cross-reactivity with those found in the antigens of other bacteria. Clearly, protein fragments containing unique epitopes represent logical alternatives to full-length protein as antigen targets in serodiagnostics as it allows for the elimination of cross-reactive epitopes, retaining only those highly specific for T. pallidum. Flagellins are usually regarded as an effective adjuvant, the significant role of which in serodiagnosis of disease has also been demonstrated in many bacteria (Robinson et al.1993; Lodes et al.2004; Wajanarogana et al.2013). Little is known about pathogenic mechanism of T. pallidum flagellin when compared to that of other bacteria. McGill et al. (2010) demonstrated that T. pallidum flagellins could react with sera from various stages of syphilis; these findings were subsequently confirmed by Jiang et al. (2016), who showed that three FlaB proteins harbor excellent sensitivity and high specificity and showed promise as diagnostic candidates for screening syphilis. Generally, the flagellin sequence conserved among pathogenic strains of target disease offers an attractive antigen candidate for serological assay. However, there also exists extensive homology between the FlaB of T. pallidum and flagellin of other bacteria (Titz et al.2006; Rajagopala et al.2007). Direct comparison of the FlaB protein sequence of T. pallidum with those of flagellin from other bacteria demonstrated high degrees of homology, especially the amino and carboxyl termini. In this study, the selected regions that show high or low homology with related regions from other bacteria were amplified by PCR and expressed as recombinant fusion proteins in E. coli. The apparent molecular weights of middle fragments were a little larger than C-terminal fragments, possibly resulting from the post-translational modification process. Unlike terminal fragments, which could react with all kinds of samples including syphilitic and negative control sera, the middle regions confined reactivity to syphilitic serum samples only. Notably, the excellent overall sensitivities for all fragments were comparable and there are no significant differences between B3M and other fragments (P > 0.05). Due to the increased level of antibody binding in healthy control sera, these cutoff values were higher for terminal fragments, reducing the sensitivity of these fragment-based ELISAs. In other words, the sensitivities of terminal fragments should be better and the specificities were actually considerably worse. Regardless of the fact that flagellin fragments located in the same region N-terminal fragments shared high levels of homology (Norris et al.1988), there are still slight differences in observed sensitivity. It is possible that the concentration of IgG directed against certain flagellin epitopes varies throughout the course of T. pallidum infection. The specificity of assays for serodiagnosis of syphilis may be improved using flagellin lacking the cross-reactive N-terminal and C-terminal. Our data showed that the three middle regions were nearly unreactive with sera from the negative control, especially the B3M, the overall specificity value of which was 100%. In contrast, the specificities for terminal fragment-based ELISAs were significantly poor due to cross-reactivity, which occurred more frequently in samples from disease control than in samples from healthy individuals. This phenomenon accorded well with the homology analysis above. Futhermore, cross-reactivity can be caused by any number of environmental, bacterial, viral or other antigens not included in the sequence databases. In particular, it is worth underlining that many samples from patients infected either with hepatitis B virus or hepatitis C virus were positively reactive with terminal fragments compared to a few with the middle regions. These samples that yielded positive reactivity to any fragments were re-screened by RPR and TPPA tests and the results were negative. We found subsequently that many of these patients had hepatocellular carcinoma when referred to medical records. This may be explained by the findings of Fedirko et al. (2017), who suggested that weakened gut barrier function and subsequent exposure to gut-derived bacterial flagellin may promote hepatocarcinogenesis. Given the fact that there are high levels of homology between FlaB protein of T. pallidum and flagellins of vast gut-derived flagellated bacteria, the larger cross-reactivity rate observed in samples with positive HBV/HCV status is not surprising. When compared to the diagnostic performances of three full-length FlaB proteins in our previous study (Jiang et al.2016), the specificity of those terminal fragments was considerably reduced. This is because the specificity differences among fragments was amplified when the whole protein was divided into several fragments. Taken together, our results clearly indicate that the cross-reactive epitopes were located in the highly conserved N- and C-terminal regions of FlaB proteins and the specific epitopes were in the middle regions. Direct comparison of the diagnostic performances of the recombinant fragment-based ELISA and the Architect Syphilis TP tests is a focus of our study. Our data showed that only B3M, the middle region of FlaB3, exhibited an equal sensitivity (95.0%) to, and a higher specificity (100%) than, this automatic chemiluminescent immunoassay (95.5% for sensitivity and 94.8% for specificity). None of the serum samples from negative controls were found to be reactive by B3M-based IgG ELISA, whereas Architect Syphilis TP exhibited positive reactivity with three samples from healthy individuals, two samples from a patient with Lyme disease, one sample from a patient with Epstein–Barr virus infection, and two samples from patients with hepatitis B virus infection. Just as speculated by some studies (Marangoni et al.2009, 2013), the source of the false positivity of Architect Syphilis TP does not seem to be cross-reactive antibodies to treponemal antigens, but is still unknown. Moreover, the overall percentage of agreement and κ value of B3M are also higher than those of Architect Syphilis TP, indicating that the B3M-based ELISA was in perfect agreement with the TPPA assay. Thus, recombinant fragment B3M may represent an effective candidate antigen for the seroscreening of syphilis. Our study had some limitations. All syphilitic serum samples were collected from Hunan, China, which is an area with a relatively high prevalence of syphilis. Therefore, effort is needed to explore the diagnostic performance of target fragment in areas with different syphilis prevalence rates and different patient populations. In addition, the serum samples tested here lack the congenital syphilis, therefore it is not suitable for evaluating the serodiagnostic value of fragment-based ELISA for early infection. Evaluations of the efficiency of antigens at all disease stages will improve accuracy. In conclusion, a novel immune-dominant FlaB fragment, B3M, has been identified that was highly sensitive for detecting different stages of syphilis and extremely specific when assayed against a large panel of negative control serum samples. Additionally, the diagnostic performance of this recombinant fragment equaled or surpassed that of either the full-length protein or Architect Syphilis TP assay with respect to sensitivity and specificity, respectively. However, before B3M can be definitively verified as helpful in the serodiagnosis of syphilis, further studies need to be conducted. Acknowledgements We thank Kanglin Wan, Chinese Centre for Disease Control and Prevention, Beijing, China; Liang Cai, Hunan Provincial Centre for Disease Control and Prevention, Changsha, China; Mingxing Liang, The First People's Hospital of Huaihua, Huaihua,China; and Xiaoping Xie, The First Affiliated Hospital, University of South China, for providing serum samples for this study. FUNDING This work was supported by the National Natural Science Foundation of China (grant numbers 81471576, 81702046, 81701577 and 81373230), the Hunan Province Cooperative Innovation Center for Molecular Target New Drug Study (2015-351), the Hunan Provincial Key Laboratory for Special Pathogens Prevention and Control Foundation (grant number 2014-5), and the General Project in Hunan Province Science and Technology Program (grant number 2014TT2025). Conflict of interest. None declared. REFERENCES Binnicker MJ. Which algorithm should be used to screen for syphilis? Curr Opin Infect Dis 2012; 25: 79– 85. Google Scholar CrossRef Search ADS PubMed Binnicker MJ, Jespersen DJ, Rollins LO. Direct comparison of the traditional and reverse syphilis screening algorithms in a population with a low prevalence of syphilis. J Clin Microbiol 2012; 50: 148– 50. Google Scholar CrossRef Search ADS PubMed Buchacz K, Klausner JD, Kerndt PR et al. HIV incidence among men diagnosed with early syphilis in Atlanta, San Francisco, and Los Angeles, 2004 to 2005. J Acquir Immune Defic Syndr 2008; 47: 234– 40. Google Scholar CrossRef Search ADS PubMed Centers for Disease Control and Prevention. Discordant results from reverse sequence syphilis screening—five laboratories, United States, 2006–2010. MMWR Morb Mortal Wkly Rep 2011; 60: 133– 7. PubMed Fedirko V, Tran HQ, Gewirtz AT et al. Exposure to bacterial products lipopolysaccharide and flagellin and hepatocellular carcinoma:a nested case-control study. BMC Medicine 2017; 15: 72. Google Scholar CrossRef Search ADS PubMed Gassmann GS, Jacobs E, Deutzmann R et al. Analysis of the Borrelia burgdorferi GeHo fla gene and antigenic characterization of its gene product. J Bacteriol 1991; 173: 1452– 9. Google Scholar CrossRef Search ADS PubMed Henao-Martinez AF, Johnson SC. Diagnostic tests for syphilis: New tests and new algorithms. Neurol Clin Pract 2014; 4: 114– 22. Google Scholar CrossRef Search ADS PubMed Hoover KW, Radolf JD. Serodiagnosis of syphilis in the recombinant era: reversal of fortune. J Infect Dis 2011; 204: 1295– 6. Google Scholar CrossRef Search ADS PubMed Jiang C, Xiao J, Xie Y et al. Evaluation of FlaB1, FlaB2, FlaB3, and Tp0463 of Treponema pallidum for serodiagnosis of syphilis. Diagn Microbiol Infect Dis 2016; 84: 105– 11. Google Scholar CrossRef Search ADS PubMed Joys TM. The covalent structure of the phase-1 flagellar filament protein of Salmonella typhimurium and its comparison with other flagellins. J Biol Chem 1985; 260: 15758– 61. Google Scholar PubMed Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977; 33: 159– 74. Google Scholar CrossRef Search ADS PubMed Lipinsky D, Schreiber L, Kopel V et al. Validation of reverse sequence screening for syphilis. J Clin Microbiol 2012; 50: 1501. Google Scholar CrossRef Search ADS PubMed Liu J, Howell JK, Bradley SD et al. Cellular architecture of Treponema pallidum: novel flagellum, periplasmic cone, and cell envelope as revealed by cryo electron tomography. J Mol Biol 2010; 403: 546– 61. Google Scholar CrossRef Search ADS PubMed Lodes MJ, Cong Y, Elson CO et al. Bacterial flagellin is a dominant antigen in Crohn disease. J Clin Invest 2004; 113: 1296– 306. Google Scholar CrossRef Search ADS PubMed Loeffelholz MJ, Binnicker MJ. It is time to use treponema-specific antibody screening tests for diagnosis of syphilis. J Clin Microbiol 2012; 50: 2– 6. Google Scholar CrossRef Search ADS PubMed Marangoni A, Moroni A, Accardo S et al. Laboratory diagnosis of syphilis with automated immunoassays. J Clin Lab Anal 2009; 23: 1– 6. Google Scholar CrossRef Search ADS PubMed McGill MA, Edmondson DG, Carroll JA et al. Characterization and serologic analysis of the Treponema pallidum proteome. Infect Immun 2010; 78: 2631– 43. Google Scholar CrossRef Search ADS PubMed Marangoni A, Nardini P, Foschi C et al. Evaluation of the BioPlex 2200 syphilis system as a first-line method of reverse-sequence screening for syphilis diagnosis. Clin Vaccine Immunol 2013; 20: 1084– 8. Google Scholar CrossRef Search ADS PubMed Mattei PL, Beachkofsky TM, Gilson RT et al. Syphilis: a reemerging infection. Am Fam Physician 2012; 86: 433– 40. Google Scholar PubMed Mizel SB, Bates JT. Flagellin as an adjuvant: cellular mechanisms and potential. J Immunol 2010; 185: 5677– 82. Google Scholar CrossRef Search ADS PubMed Morshed MG, Singh AE. Recent trends in the serologic diagnosis of syphilis. Clin Vaccine Immunol 2015; 22: 137– 47. Google Scholar CrossRef Search ADS PubMed Nguyen CT, Hong SH, Sin JI et al. Flagellin enhances tumor-specific CD8+ T cell immune responses through TLR5 stimulation in a therapeutic cancer vaccine model. Vaccine 2013; 31: 3879– 87. Google Scholar CrossRef Search ADS PubMed Nguyen CT, Kim SY, Kim MS et al. Intranasal immunization with recombinant PspA fused with a flagellin enhances cross-protective immunity against Streptococcus pneumoniae infection in mice. Vaccine 2011; 29: 5731– 9. Google Scholar CrossRef Search ADS PubMed Norris SJ; Treponema pallidum Polypeptide Research Group. Polypeptides of Treponema pallidum: progress toward understanding their structural, functional, and immunologic roles. Microbiol Rev 1993; 57: 750– 79. Google Scholar PubMed Norris SJ, Charon NW, Cook RG et al. Antigenic relatedness and N-terminal sequence homology define two classes of periplasmic flagellar proteins of Treponema pallidum subsp. pallidum and Treponema phagedenis. J Bacteriol 1988; 170: 4072– 82. Google Scholar CrossRef Search ADS PubMed Ong D, Bhardwaj G, Ong J et al. Keeping an eye on syphilis. Aust Fam Physician 2017; 46: 401– 4. Google Scholar PubMed Rajagopala SV, Titz B, Goll J et al. The protein network of bacterial motility. Mol Syst Biol 2007; 3: 128. Google Scholar CrossRef Search ADS PubMed Robinson JM, Pilot-Matias TJ, Pratt SD et al. Analysis of the humoral response to the flagellin protein of Borrelia burgdorferi: cloning of regions capable of differentiating Lyme disease from syphilis. J Clin Microbiol 1993; 31: 629– 35. Google Scholar PubMed Samoff E, Koumans EH, Gibson JJ et al. Pre-treatment syphilis titers: distribution and evaluation of their use to distinguish early from late latent syphilis and to prioritize contact investigations. Sex Transm Dis 2009; 36: 789– 93. Google Scholar CrossRef Search ADS PubMed Smith BC, Simpson Y, Morshed MG et al. New proteins for a new perspective on syphilis diagnosis. J Clin Microbiol 2013; 51: 105– 11. Google Scholar CrossRef Search ADS PubMed Stamm LV. Syphilis: Re-emergence of an old foe. Microb Cell 2016; 3: 363– 70. Google Scholar CrossRef Search ADS PubMed Stamm LV, Drapp RL. A synthetic lymph node containing inactivated Treponema pallidum cells elicits strong, antigen-specific humoral and cellular immune responses in mice. Pathog Dis 2014; 70: 88– 94. Google Scholar CrossRef Search ADS PubMed Titz B, Rajagopala SV, Ester C et al. Novel conserved assembly factor of the bacterial flagellum. J Bacteriol 2006; 188: 7700– 6. Google Scholar CrossRef Search ADS PubMed Tsang RS, Martin IE, Lau A et al. Serological diagnosis of syphilis: comparison of the Trep-Chek IgG enzyme immunoassay with other screening and confirmatory tests. FEMS Immunol Med Microbiol 2007; 51: 118– 24. Google Scholar CrossRef Search ADS PubMed Wajanarogana S, Nimnuch P, Thongmee A et al. Potential of recombinant flagellin fragment from Burkholderia thailandensis as an antigen for melioidosis antibody detection by indirect ELISA. Mol Cell Probes 2013; 27: 98– 102. Google Scholar CrossRef Search ADS PubMed Woolston SL, Dhanireddy S, Marrazzo J. Ocular syphilis: a clinical review. Curr Infect Dis Rep 2016; 18: 36. Google Scholar CrossRef Search ADS PubMed World Health Organization. Prevalence and incidence of selected sexually transmitted infections: Chlamydia trachomatis, Neisseria gonorrhoeae, syphilis and Trichomonas vaginalis: methods and results used by WHO to generate 2005 estimates . http://apps.who.int/iris/bitstream/10665/44735/1/9789241502450_eng.pdf. World Health Organization 2011 (19 January 2018, date last accessed). Xu M, Xie Y, Jiang C et al. A novel ELISA using a recombinant outer membrane protein, rTp0663, as the antigen for serological diagnosis of syphilis. Int J Infect Dis 2016; 43: 51– 7. Google Scholar CrossRef Search ADS PubMed Zgair AK, Chhibber S. Stenotrophomonas maltophilia flagellin restricts bacterial colonization in BALB/c mouse lung in vivo. FEMS Immunol Med Microbiol 2012; 66: 191– 200. Google Scholar CrossRef Search ADS PubMed © FEMS 2017. All rights reserved. For permissions, please e-mail:firstname.lastname@example.org
Pathogens and Disease – Oxford University Press
Published: Mar 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.
All for just $49/month
Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.
Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.
It’s easy to organize your research with our built-in tools.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera