The T. pallidum Nichols strain was propagated in healthy male NZW rabbits by intratesticular inoculation and harvested as previously described by Lukehart et al. (33). All rabbits were screened by rapid plasma reagin test (RPR) and Treponema pallidum particle agglutination assay (TPPA) to exclude T. paraluiscuniculi infection.

For expression of rTp0954 protein in Escherichia coli, tp0954 gene (GenBank accession number AAC65909.1, 1437-bp full length) coding full-length Tp0954 protein with 478 amino acid residues without the N-terminal signal peptide (19 amino acid residues) and with hexahistidine tag sequence, was PCR amplified from T. pallidum (Nichols strain) genomic DNA using the Sense: 5′-ATGGGATCCTGTGTTAGCACCGGTTCTAACTCTG-3′ and Antisense: 5′-CATCTCGAGTTACGGGTTGGTACGCAGGGA-3′ primers with BamHI and XhoI restriction sites underlined. The amplicon was cloned into pET30a to produce the recombinant expression plasmid pET30a/tp0954 whose sequence was confirmed by sequencing. The tp0954 construct was transformed into the expression strain BL21 Star E. coli (Invitrogen, Carlsbad, CA, USA) and rTp0954 expression was induced with 0.5 mM isopropyl-β-d-thiogalactopyranoside (IPTG) for 4 h at 25°C, followed by purification with Ni-NTA columns (Qiagen, Hilden, Germany) and treatment with Detoxi-Gel™ Endotoxin Removing Gel (Pierce Biotechnology, Inc., Rockford, IL, USA) to remove endotoxin. A bicinchoninic acid (BCA) protein assay kit (CWBIO, Beijing, China) was used to determine the protein concentrations. The target proteins were separated by SDS-PAGE and transferred to a polyvinylidene fluoride (PVDF) membrane (Merck Millipore, Darmstadt, Germany). Using 6×-His Tag Mouse Monoclonal Antibody (1:1000; Proteintech, Rosemount, IL, USA) or T. pallidum-infected rabbit serum (1:1000; the Institute of Pathogen Biology at the University of South China) as the primary antibody, and horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG (1:10,000, Proteintech) or goat anti-rabbit IgG (1:10,000, Proteintech) as the secondary antibody, respectively, protein bands were analyzed using the eECL Western Blot Kit (CWBIO) with a G:BOX Chemi XX9 (Syngene, Cambridge, UK) digital imager.

A cohort of 18 RPR and TPPA-seronegative male NZW rabbits (8-13 weeks old, weighing about 3.0 kg; the Department of Experimental Zoology of the University of South China, Hengyang, China) were used for immunization. The animals were randomly selected into 3 groups: PBS-immunized control group (N = 5), Freund’s adjuvant (FA)-immunized control group (N = 5), and rTp0954-Freund’s adjuvant (rTp0954/FA)-immunized group (N = 8). For the rTp0954/FA-immunized group, each rabbit was sedated and immunized once every two weeks for three times with 250 μg purified rTp0954 protein (1 mg/ml) mixed with an equal volume of Freund’s adjuvant system (Sigma-Aldrich, St. Louis, MO, USA; complete Freund’s adjuvant for the first immunization and incomplete Freund’s adjuvant for the booster immunizations) at each immunization. Each immunization was delivered equally as 4 subcutaneous (0.1 ml per site) injections in the necks and 2 intramuscular (0.05 ml per site) injections in the quadricep muscles. FA-immunized and PBS controls received PBS with the FA system in a 1:1 ratio or PBS alone at corresponding time points, respectively.

Two weeks after the final immunization, 11 randomly selected rabbits (N = 3, PBS control; N = 3, FA-immunized; N = 5, rTp0954/FA-immunized) were sedated, back-shaved, and cleaned with 75% ethanol. Each rabbit was challenged intradermally with 0.1 ml of 1×10⁶ freshly isolated T. pallidum Nichols strain per ml in 0.9% saline at each of 8 sites (each rabbit received 0.8×10⁶ T. pallidum). The development of lesions, including erythema, induration and painless ulceration, was monitored and then the lesion diameters were determined every 2 days.

At weeks 0, 2, 4, and 6 post the first immunization and prior to intradermal challenge with T. pallidum, blood samples were collected from ear veins of all rabbits to obtain the serum for the detection of rTp0954-specific antibody titers by ELISA. Briefly, the 96-well plates were coated with 1 μg purified rTp0954 and blocked with 5% non-fat milk in PBST. Rabbit serum was two-fold serially diluted starting from a dilution of 1:100. HRP-conjugated goat anti-rabbit IgG (1:10,000, Proteintech) was used as a secondary antibody, and binding was detected by using TMB (Tetramethylbenzidine) peroxidase substrate. After addition of stop solution, the absorbance at 450 nm (A₄₅₀ value) was detected with a microplate reader (BioTek Instruments, Winooski, VT, USA). Each experiment was repeated three times.

Two weeks after the final immunization, 7 randomly selected rabbits (N = 2, PBS control; N = 2, FA-immunized; N = 3, rTp0954/FA-immunized) were euthanized and spleens were removed for preparation of splenic cell suspension as previously described (23). Splenocytes were resuspended in RPMI-1640 medium (Sigma Aldrich, St. Louis, MI, USA) containing 10% fetal bovine serum (FBS; Biological Industries, Beit Haemek, Israel), penicillin (100 U/ml) and streptomycin (100 μg/ml) for the following detection of lymphocyte proliferation and secretion level of IFN-γ from splenocytes.

The proliferation of lymphocytes was detected by using the Enhanced Cell Counting Kit-8 (CCK-8; Beyotime, Shanghai, China) according to the manufacturer’s instructions. Briefly, rabbit splenocyte suspension at 1×10⁶ cells/well in 200 μl was stimulated with 5 μg/ml of rTp0954. In addition, splenocyte suspension cultures with 5 μg/ml of concanavalin A (Con A, Sigma Aldrich) and PBS stimulation were used as a positive control and blank control, respectively. Following the incubation for 48 h, 20 μl of CCK-8 solution was then added to each well for 4 h prior to the end of culture. The A₄₅₀ value was detected with a microplate reader (BioTek Instruments). Independent experiments were conducted in triplicate. A₄₅₀ value was used to express the results, and the stimulation index (SI) was calculated according to the following formula: SI(%)=At/Ac×100% (At and Ac represents the mean A₄₅₀ value of the experimental group and the blank control group, respectively).

IFN-γ was assessed in this study because it is an indicator of Th1 response. The rabbit splenic lymphocytes collected above were cultured in 1-cm-diameter-well culture plates (Costar, Cambridge, Mass., USA) plates with at 6×10⁵ cells in 500 μl. Each well received 5 μg/ml of rTp0954 protein and PBS as blank control, followed by incubation at 37°C for 72 h. Supernatants from cultured cells were harvested for detection of the level of IFN-γ secretion in splenocytes according to the instruction of the Rabbit IFN-γ ELISA Flex kit (Mabtech, Nacka Strand, Sweden). Independent experiments were conducted in triplicate. The A₄₅₀ value was measured using a microplate reader (BioTek Instruments).

DNAs were extracted from tissues of T. pallidum-infected rabbits, including cutaneous lesions, blood, liver and spleen, by using a TIANamp Genomic DNA Kit (Tiangen, Beijing, China) following the manufacturer’s protocol. 3 tissue specimens were taken from each animal organ to ensure good reproducibility. PCR amplification was performed by using a LightCycle 96 apparatus (Roche, Basel, Switzerland). Quantitative analysis of T. pallidum gDNA and rabbit gDNA was achieved by using primers targeting T. pallidum endoflagellar sheath protein (flaA) gene and rabbit collagenase-1 precursor (MMP-1) gene, respectively. The sequences of the primers are given as followings: flaA Sense: 5’-GCGGTTGCACAGTGGGAG-3’, Antisense: 5’-CAGCATGGGCGACAGGAT-3’; MMP-1 Sense: 5’-TTGCTTCTTCACACCAGAATGCTGT-3’, Antisense: 5’-GCGTGATCAGGCACTATGTAGCAAT-3’. The primers for quantitative real-time PCR (qPCR) amplification were provided by Sangon Biotech (Shanghai, China). As directed by the manufacturer, qPCR amplifications were performed in 20-μl SuperReal PreMix Plus Kit (SYBR Green) (Tiangen, Beijing, China) reaction mixture. Standard curves were generated for flaA and MMP-1 by using a 10-fold dilutions in series from 10⁷ to 10¹ reproductions of T. pallidum gDNA and 2-fold dilutions in series of rabbit gDNA from 100 to 0.0488 ng/μl. The efficiency was over 99%, and the R2 value was above 0.99 in every tests, respectively. The amount of gDNA in different tissues was normalized according to the lowest gDNA concentration determined by the spectrophotometric measurements. The following conditions were used for the PCR of flaA and MMP-1: 95°C for 15 min (flaA) or 10 min (MMP-1), 40 cycles of 95°C for 10 sec, 58°C (flaA) or 55°C (MMP-1) for 20 sec, and 72°C for 20 sec. Following is an analysis of melt-curves: 95°C for 10 sec, 65°C for 60 sec, and 97°C for 1 sec. Repeat each sample three times and with a no-template control.

On day 21 post-challenge, each rabbit skin lesion was removed and subdivided using a 4-mm punch. The tissues taken from cutaneous lesions were fixed, embedded and sectioned for histopathologic analysis. After conventional staining with hematoxylin and eosin (H&E), the degree of inflammatory cell infiltration in each tissue section was observed and analyzed under a microscope.

On day 21 post-challenge, after euthanasia, the PLNs from randomly selected 2 rTp0954-immunized rabbits and 2 PBS control rabbits were removed and disrupted using a 70-μm cell sieve in a 6-well sterile plate with 0.9% saline. Four naïve anesthetized NZW rabbits (negative RPR and TPPA) were randomly selected and the strained suspension was injected into their testis with 1 ml per testis, as described previously by Lukehart et al. (33). From the 7^th day, signs of orchitis of rabbits were monitored every day and seroconversion (RPR and TPPA) were assayed every 7 to 10 days. The animals that showed signs of orchitis and seroconversion (positive TPPA and 1:8 RPR titer) or that that did not develop orchitis or seroconversion until day 100 were euthanized, and the testes were harvested. T. pallidum in testicular exudates were purified, concentrated (33) and visualized by silver staining under light microscope.

Statistical analysis was performed with GraphPad Prism 8.0 software. Brown–Forsythe and Welch ANOVA test were used to evaluate the significance of statistical differences between the groups. Differences in T. pallidum load among tissues were analyzed by Ordinary one-way ANOVA and Dunnett test (P < 0.05 indicated a significant difference and results are reported as Mean ± SD).

The Tp0954 gene was amplified from the T. pallidum (Nichols strain) genomic DNA ( Figure 1A ). The soluble rTp0954 protein containing the His epitope tag with approximate 55-kDa molecular mass was expressed and determined by SDS-PAGE analysis with an estimated purity of 95% ( Figure 1B ). Western blot analysis showed that target protein was only reactive with anti-His monoclonal antibody and T. pallidum-infected rabbit sera, not with normal rabbit sera (NRS) ( Figure 1C ). ELISA results showed that, beginning on week 2, rTp0954 induced gradually increasing titers of anti-Tp0954 IgG antibodies compared with PBS or FA-immunized control rabbits ( Figure 2 ), indicating that immunization of rabbits with rTp0954 generates a specific antibody response.

Two weeks after the final immunization, rabbit splenic lymphocytes were isolated from different groups and their proliferative responses specific to rTp0954 were measured by CCK-8 assay, with Con A and PBS stimulation as a positive and blank control, respectively. As shown in Figure 3 , the positive (Con A) control showed the highest SI values, and the SI values in rTp0954-immunized groups were significantly higher than those in PBS- or FA-immunized groups (P < 0.01), suggesting that rTp0954 immunization induces significant proliferation of splenic lymphocytes.

IFN-γ is a critical Th1-related cytokine, which is essential for early clearance of T. pallidum in lesions. The levels of IFN-γ produced by rabbit splenic cells were measured by ELISA. As shown in Figure 4 , IFN-γ levels in rTp0954-immunized groups were significantly increased when compared with those in PBS- or FA-immunized groups (P < 0.0001), which closely resembles the proliferative response of splenic lymphocytes specific to rTp0954.

2 weeks post the final immunization, each animal was infected at 8 sites on the back with a total of 0.8 × 10⁶ of freshly isolated T. pallidum. Since the painless indurated lesion (chancre), that develops at the site of initial infection is the first clinical sign characteristic of early syphilis, monitoring the development of chancre provides a measure of the effectiveness of syphilis vaccine. During the subsequent 21 days, lesions were monitored daily for diameter, induration and ulceration. The observed results showed that rTp0954/FA-immunized animals exhibited delayed lesion development. On day 9, there were 21 and 20 lesion indurations (a total of 24) observed in FA-immunized and PBS control animals, respectively. However, by day 14, only 24 lesion indurations (total of 40) were observed in animals immunized with rTp0954/FA ( Table 1 ). Moreover, on day 21 post-challenge, the incidence of ulcerative lesions in rTp0954/FA-immunized animals (12.5%) was much lower than that in PBS control (66.7%) and FA control animals (60.0%) ( Figures 5A–C ; Table 1 ). Additionally, over the 21-day measurement period post-challenge, the diameter of cutaneous lesions in the rTp0954-immunized animals was consistently smaller than that in the PBS and/or FA control animals ( Figure 5D ). During the measurement period of 9-21 days, the mean lesion size in the rTp0954-immunized animals was much smaller than that in the PBS and FA control animals( Table 1 ). All these results indicate that immunization of rabbits with rTp0954 is able to significantly delay lesion development and decrease the diameter of cutaneous lesions.

To evaluate the ability of rTp0954 immunization to prevent T. pallidum dissemination, all animals challenged with mobile T. pallidum-were euthanized on day 21 post-challenge. qPCR was performed to determine T. pallidum DNA concentration to assess the treponemal burdens in biopsy specimens obtained from cutaneous lesions at the initial infection sites as well as distal tissues or organs, including blood, spleen and liver. The results revealed that there was a noticeable decline in the T. pallidum DNA concentration at the initial infection sites in the rTp0954/FA-immunized animals compared to the FA and PBS controls (P < 0.0001) ( Figure 6A ). Owing to highly invasive ability of T. pallidum that spreads through the blood to various distal tissues and organs where treponemes could be detected (15, 19, 28, 34), the treponemal burdens in the blood, liver, and spleen of the animals were also assessed. As shown in Figures 6B–D , the treponemal burdens in the blood, livers, spleens from rTp0954-immunized rabbits were still obviously decreased than that in the FA and PBS controls (P < 0.01, 0.001 or 0.001), suggesting that immunization of rabbits with rTp0954 prevents treponemal dissemination to distant organs.

In order to further determine the ability of rTp0954 immunization to prevent treponemal dissemination, PLNs from 2 PBS control and 2 Tp0954/FA-immunized animals were transferred into 4 naïve recipient animals. The infection status of the rabbits receiving PLNs was determined by orchitis development, seroconversion (RPR and TPPA) and microscopic observation of silver stained-treponemes in testicular aspirates. Two PBS control naïve recipient rabbits (R-Ct1 and R-Ct2) developed overt orchitis and stronger seroconversion (1:8 RPR titer and positive TPPA) on day 40 and day 56 post transfer ( Table S1 ), respectively. Silver-stained treponemes under microscopy in the concentrated testicular extracts from R-Ct1 and R-Ct2 also were observed ( Figure S1A ). Conversely, two naïve PLN recipients from the Tp0954/FA-immunized rabbits (R-Im1 and R-Im2) developed neither orchitis nor seroconversion ( Table S1 ). Moreover, no treponemes were observed in testicular exudates by silver staining 100 days after transfer ( Figure S1B ). All the findings suggest that rTp0954 vaccination prevents T. pallidum dissemination in rabbits.

On day 21 post-challenge, H&E stained sections from primary cutaneous lesion sites of rabbits were performed to evaluate the intensities of inflammatory infiltration and cell types. All lesions revealed various degrees of infiltrates of lymphoplasmacytic cells (mainly including lymphocytes, plasma cells, and macrophages), characteristic of syphilis (35), with a significant increase in the degree of inflammatory infiltration in primary lesions from rTp0954/FA-immunized animals when compared with PBS or FA controls ( Figures 7A–F ). Consistent with this observation, rTp0954/FA-immunized rabbits had the lowest levels of treponemal burdens in the primary skin challenge sites and the distal disseminated infection sites, including blood, liver and spleen.