Genetic engineering of Treponema pallidum subsp. pallidum, the Syphilis Spirochete

Despite more than a century of research, genetic manipulation of Treponema pallidum subsp. pallidum (T. pallidum), the causative agent of syphilis, has not been successful. The lack of genetic engineering tools has severely limited understanding of the mechanisms behind T. pallidum success as a pathogen. A recently described method for in vitro cultivation of T. pallidum, however, has made it possible to experiment with transformation and selection protocols in this pathogen. Here, we describe an approach that successfully replaced the tprA (tp0009) pseudogene in the SS14 T. pallidum strain with a kanamycin resistance (kanR) cassette. A suicide vector was constructed using the pUC57 plasmid backbone. In the vector, the kanR gene was cloned downstream of the tp0574 gene promoter. The tp0574prom-kanR cassette was then placed between two 1-kbp homology arms identical to the sequences upstream and downstream of the tprA pseudogene. To induce homologous recombination and integration of the kanR cassette into the T. pallidum chromosome, in vitro-cultured SS14 strain spirochetes were exposed to the engineered vector in a CaCl2-based transformation buffer and let recover for 24 hours before adding kanamycin-containing selective media. Integration of the kanR cassette was demonstrated by qualitative PCR, droplet digital PCR (ddPCR), and whole-genome sequencing (WGS) of transformed treponemes propagated in vitro and/or in vivo. ddPCR analysis of RNA and mass spectrometry confirmed expression of the kanR message and protein in treponemes propagated in vitro. Moreover, tprA knockout (tprAko-SS14) treponemes grew in kanamycin concentrations that were 64 times higher than the MIC for the wild-type SS14 (wt-SS14) strain and in infected rabbits treated with kanamycin. We demonstrated that genetic manipulation of T. pallidum is attainable. This discovery will allow the application of functional genetics techniques to study syphilis pathogenesis and improve syphilis vaccine development.


Introduction
Syphilis is a chronic sexually transmitted infection that still represents a significant burden for public health as it causes significant morbidity and mortality worldwide. The World Health Organization (WHO) estimates that syphilis global incidence ranges between 5.6 to 11 million new cases every year, while disease prevalence is between 18 to 36 million cases worldwide [1,2]. Although most of those cases occur in low-and middle-income countries where the disease is endemic, syphilis rates have been steadily increasing in high-income countries for decades now, including in the US. In these countries, mainly men who have sex with men (MSM) and persons living with HIV (PLHIV) are affected [3][4][5][6][7][8]. In the US, the rate of early syphilis in 2019 (11.9 cases per 100,000 population), represented a 460% increase compared to the cases reported in 2000 (2.1 cases per 100,000 population) [3]. If untreated, syphilis can progress to affect the patient's cardiovascular and central nervous systems, possibly leading to serious manifestations such as aortic aneurism, stroke, hearing or visual loss, dementia, and paralysis [9]. Furthermore, mother-to-child transmission of syphilis during pregnancy accounts for up to 50% of stillbirths in sub-Saharan Africa and a high proportion of perinatal morbidity and mortality cases [10].
A better understanding of T. pallidum biology and syphilis pathogenesis would help in devising more effective measures for disease control. Recently, Edmondson et al. [11][12][13] described a method to continually propagate T. pallidum in vitro using a cell culture-based system previously pioneered by Fieldsteel et al. [14]. This method represented a major advancement in the field, in that it provided investigators with an alternative to the propagation of treponemal strains in laboratory rabbits. Despite such advancement, a limitation in the study of T. pallidum remained the lack of tools for genetic modification of this pathogen. The availability of the cultivation system, however, paved the way to experimenting with transformation and selection procedures to introduce foreign DNA into T. pallidum genome.
Here, we describe a protocol that successfully allowed us to integrate a kanamycin resistance (kan R ) cassette into a pseudogene (tprA, encoded by the tp0009 gene) of the T. pallidum SS14 strain using a suicide vector carrying the resistance gene between two homology arms identical to the regions upstream and downstream of the pseudogene. The confidence in our achievement is supported by the results of complementary techniques including qualitative PCR and RT-PCR, quantitative droplet digital PCR (ddPCR) and RT-ddPCR, whole genome sequencing (WGS), kanamycin susceptibility assay, mass spectrometry (MS), and in vivo propagation of transformed treponemes.
Although we ablated a pseudogene, which did not allow us to obtain a mutant lacking a known phenotype to be evaluated through functional assays, our work provides evidence that transformation and genetic manipulation of T. pallidum is attainable. This is a significant step forward in the field. Our discovery opens numerous possibilities, including classical genetic studies in this pathogen, the long-awaited application of functional genomics techniques, and even the possibility of targeting virulence factors responsible for immune evasion and persistence to obtain an attenuated strain to inform vaccine development efforts.

Plasmid construction
The transformation plasmid was a pUC57-based suicide vector where the kan R gene was placed between two~1 kbp homology arms identical to the regions upstream (998 bp) and downstream (999 bp) of the tprA frameshifted ORF, respectively. To drive expression of the kan R gene, the promoter of the tp0574 gene (encoding the 47 kDa lipoprotein), previously identified by Weigel et al. [15], was chosen based on experimental evidence that tp0574 is among the most highly transcribed genes in T. pallidum, and is possibly expressed constitutively in this pathogen [16,17]. The kan R gene sequence was derived from the Rts1 plasmid, a large conjugative plasmid originally isolated from Proteus vulgaris [18]. Codon optimization was not performed because the GC content of the kan R gene (44%) was deemed to be sufficiently close to the T. pallidum genome GC content (~51%).

Transformation and selection of T. pallidum
The pUC57-based ptprAarms-tp0574prom-kan R plasmid construct was used to transform wild-type SS14 (wt-SS14) treponemes. A flowchart of the transformation procedure is provided in Fig 1A. Transformed T. pallidum cells were subsequently propagated in kanamycinsupplemented media (25 μg/ml, effective in selection experiments of other spirochetes). Because the transformation buffer contained CaCl 2 to increase membrane permeability and facilitate plasmid intake, we also exposed wt-SS14 cells to transformation buffer alone (without plasmid, to exclude CaCl 2 lethality for the treponemes) and proceeded to propagate these cells in culture media with no antibiotic. Furthermore, to ensure that non-transformed treponemes would not survive exposure to kanamycin at the concentration used for in vitro selection, the wt-SS14 strain was also propagated in media containing 25 μg/ml of kanamycin. Due to the long generation time of the SS14 strain (~44 hours), T. pallidum cells were sub-cultured every 14 days instead of every week until Passage #7 (Week 12 post-transformation; Fig 1B), and weekly thereafter. Transformed tprA knockout (tprA ko -SS14) treponemes could be microscopically counted 2 weeks post-transformation (Fig 1B; Passage #2), even though only 2.4 treponemes per dark-field microscope (DFM) field could be seen at this time (corresponding to a concentration of 2.4x10 6 T. pallidum cells/ml). For transformed treponemes, cell density increased four weeks post-transformation and remained steady throughout Passage #6 (Week 10 post-transformation). During this window (Passage #2-6), the average number of treponemes counted was 2.8x10 7 cells/ml. The density of wt-SS14 treated with CaCl 2 alone was higher than that of tprA ko -SS14 cells already at Passage #2 (Week 2 post-exposure; Fig 1B), suggesting that treponemes were not harmed by CaCl 2 . Propagation of this strain was halted at Passage #6 (Week 10 post-exposure; Fig 1B). Wild-type SS14 cells propagated in kanamycincontaining media could not be seen on the DFM over 10 weeks of propagation, confirming the ability of 25 μg/ml kanamycin to inhibit T. pallidum growth in vitro. During Passage #7-10, the tprA ko -SS14 treponemal inoculum was increased to obtain enough cells for subsequent experiments. During these passages, the average number of treponemal cells counted was 2.8x10 8 cells/ml. Overall, these data support that a kanamycin-resistant strain was obtained due to the transformation of the wt-SS14 strain with the ptprAarms-tp0574prom-kan R vector.
Qualitative PCR and quantitative ddPCR to confirm integration of the kan R gene and qualitative RT-PCR to evaluate kan R gene expression At Passage #8 (Week 13 post-transformation; Fig 1B), tprA ko -SS14 cells were harvested and processed for a) PCR to confirm integration of the kan R gene into the tprA locus, b) RT-PCR

PLOS PATHOGENS
First successful genetic manipulation of T. pallidum to assess the presence of message for the kan R gene, and c) for quantitative ddPCR to evaluate the ratio between the kan R gene and three other targets: tprA (tp0009), dnaA (tp0001), and tp0574. Samples from the wt-SS14 propagated in parallel to the tprA ko -SS14 strain were used as control. A qualitative, long-range PCR approach was first used to confirm integration of the tp0574prom-kan R sequence into the tprA locus, according to the schematic reported in Fig 2A. To this end, primers annealing to the T. pallidum genomic region flanking the tprA homology arms of the construct and to the kan R gene, respectively, were employed. In these reactions, DNA samples extracted from tprA ko -SS14 cells as well as from wt-SS14 were amplified with primer pairs 1+2, 3+4, 1+3, 5+6, and 2+4 (Fig 2A; primer sequences in Table 1). Amplification of the tp0574 gene was performed as a positive control. As expected, amplification using primer pairs 1+2 and 3+4 (Fig 2B, sub-panels a and b, respectively) yielded a positive result only when the tprA ko -SS14 DNA was used as the template, showing that integration of the kan R gene occurred. Amplification of the kan R gene was positive only from DNA extracted from the tprA ko -SS14, and the transformation plasmid DNA, used as positive control (subpanel c), while negative amplification using the 5+6 primer pair (annealing to the vector backbone; sub-panel d) showed no residual plasmid in the tprA ko -SS14 culture and that the kan R amplicon (in sub-panel c) was not due to residual vector used for transformation weeks earlier. When used together, primers 2+4 generated a 3,746 bp amplicon with the tprA ko -SS14 strain DNA template (sub-panel e), which was the expected size if the small kan R gene (816 bp in size) replaced the tprA pseudogene (1,821 bp), and a 4,643 bp amplicon in the wild-type and undetectable residual wt-SS14 in the tprA ko -SS14 culture wells. As expected, the amplification of the tp0574 gene was uniformly positive for both the transformed and wild-type strain (subpanel f). Overall, these data showed that the kan R gene was integrated into the tprA ko -SS14 strain genome in place of the tprA pseudogene and no residual plasmid could be amplified from the culture.

PLOS PATHOGENS
First successful genetic manipulation of T. pallidum cDNA was used as template to assess transcription of the kan R gene and tp0574 genes, as well as the tprA gene. Previous studies on other T. pallidum strains with a tprA pseudogene suggested that this locus is transcribed at a very low level in these strains, even though the coding sequence contains a frameshift that would truncate the resulting peptide during translation [19]. Results ( Fig 2B) showed that the kan R gene is expressed only in the tprA ko -SS14 (subpanel g). As expected, tp0574 was expressed in both strains (sub-panel h). tprA-specific mRNA could not be detected in either sample from the tprA ko -and wt-cultures harvested at this time point; however, in samples harvested in subsequent passages, tprA-specific message was detected in wt-SS14 propagated alongside the tprA ko -SS14 strain, but never in the tprA ko -SS14 strain. These results supported that the kan R transgene is actively transcribed from the tp0574 promoter in the tprA ko -SS14 strain while the tprA pseudogene is no longer transcribed in this strain due to tprA ablation. We next performed droplet digital PCR (ddPCR) on kan R , dnaA, and tprA loci in a separate laboratory to evaluate copy number ratios among these genes. In the tprA ko -SS14 strain, the kan R :dnaA ratio was equal to 1.05, while the tprA:dnaA ratio was virtually zero (0.006; Fig 3). These data reiterated that a) integration of the kan R gene occurred in the tprA locus, b) that this replacement was stable, and c) that no extra copies of the kan R gene existed outside of the T. pallidum genome. On the contrary, when the wt-SS14 DNA was used as template, the tprA: dnaA was 1.01, while the kan R :dnaA ratio was also virtually zero (0.005; Fig 3).

Whole-genome sequencing from in vitro-propagated tprA ko -SS14 treponemes
At Passage #9 (Week 14 post-transformation; Fig 1B), treponemes were harvested for wholegenome sequencing (WGS) using a custom hybridization capture panel to enrich for T. pallidum to demonstrate integration of the kan R gene into the tprA locus. Genome sequencing showed the absence of the tprA ORF in the tprA ko -SS14 strain when read assembly was performed using the wt-SS14 genome ( Fig 4A). However, when assembly used a template genome where tprA was replaced with the kan R gene, results showed replacement of tprA with the tp0574promoter-kan R sequence ( Fig 4B). The wild-type SS14 propagated in parallel to the knockout strain was also sequenced as a control. In the wt-SS14, the tprA locus is intact (Fig 4C), while a gap appeared when reads from the wt-SS14 strain were assembled to the tprA ko -SS14 strain genome with kan R in place of tprA ( Fig 4D). Because our sequencing approach used enrichment probes based on multiple wild-type T. pallidum genomes and did not include probes for the kan R gene, coverage of the kan R gene (Fig 4B) was slightly lower than the average for the other regions of the T. pallidum genome ( Fig 4B). These results showed again that the kan R gene replaced the tprA locus in the tprA ko -SS14 strain. A search for reads matching to the plasmid backbone was also conducted but yielded no mapping reads, showing lack of residual plasmid in the culture.

Kanamycin susceptibility assay
tprA ko -SS14 were grown in 25 μg/ml of kanamycin during routine propagation, a concentration shown to be treponemicidal for the wild-type (Fig 1B). To demonstrate that the tprA ko -SS14 strain could grow at significantly higher kanamycin concentration, tprA ko -SS14 cells harvested at Passage #10 (Week 15 post-transformation; Fig 1B) were used to further assess resistance to kanamycin by performing an in vitro susceptibility assay. To this end, we grew both the wt-SS14 and tprA ko -SS14 strains in media supplemented with kanamycin ranging from 200 to 1.6 μg/ml, for a total of 8 different concentrations tested in 8 replicate wells. Quantification of treponemal concentration, measured by qPCR targeting the tp0574 gene showed that media supplemented with 200 down to an MIC of 3.13 μg/ml of kanamycin strongly inhibited the growth of the wildtype strain when compared to no-antibiotic wells (Fig 5A), but did not affect the tprA ko -SS14 strain, which grew as if no antibiotic was added (Fig 5B), thus confirming that tprA ko -SS14 treponemes had become resistant to kanamycin. Furthermore, the tprA ko -SS14 strain also appeared

PLOS PATHOGENS
First successful genetic manipulation of T. pallidum to have a growth advantage compared to the wild-type strain. When growth was compared at day 7 post-inoculation using the no-antibiotic wells, the tprA ko -SS14 was shown to have grown significantly faster than the wild-type, with an average of~12,000 genome copies/μl, compared to the~4,300 copies of the wild-type strain (p<0.05), even though the initial inoculum size was the same. To ensure no residual plasmid was present in these cultures, DNA extracted from the eight tprA ko -SS14 replicate cultures grown in 200 μg/ml of kanamycin was amplified using primers specific for the backbone of the pUC57 plasmid (pair #5/6, as shown in Fig 2A). All these amplifications yielded a negative result unless the ptprAarms-tp0574prom-kan R construct was used as positive control (Fig 5C), confirming no presence of residual plasmid in these cultures. When the same DNA was tested with primers flanking the tprA homology arms, the tprA ko -SS14 cultures yielded a~3.7Kbp amplicon, expected due to the replacement of the tprA locus with the shorter kan R sequence, while the wt-SS14 strain DNA yielded a~4.6 Kbp amplicon ( Fig 5D).
The ratio kan R :dnaA in the tprA ko -SS14 cultures grown at different kanamycin concentrations estimated by ddPCR ranged between 1.07 and 1.14 ( Fig 6A) on average, while in wt-SS14 the kan R :dnaA ratio was zero. In the tprA ko -SS14 cultures, the tprA:dnaA ratio was virtually zero (0.004; Fig 6B), while the tprA:dnaA ratio was 1.18 on average ( Fig 6B). The ratio kan R : tp0574 in tprA ko -SS14 grown in different kanamycin concentrations, was shown to be on average 1.40 in treponemes analyzed after 7 days in culture, and slightly higher (1.66) in  Table 1), showing that in the tprA ko -SS14 strain a smaller amplicon is obtained due to the replacement of tprA by the kan R gene, which is approximately 1Kb shorter in size. Wild-type SS14 DNA template yielded a longer amplicon due to an intact tprA locus still in place. No amplification was detected using the transformation plasmid DNA control. NTC: no template control; W: wt-SS14 strain DNA; P: ptprAarms-tp0574prom-kan R plasmid DNA template; M: molecular size marker (Kbp). https://doi.org/10.1371/journal.ppat.1009612.g005

PLOS PATHOGENS
First successful genetic manipulation of T. pallidum treponemes harvested after 4 days in culture (Fig 6C), although this difference was not significantly different. This result suggested that more copies of the kan R gene than the tp0574 gene were present in the extracted DNA at sample harvest. This was overall an expected result. The replacement of the tprA pseudogene (tp0009) in the SS14 strain, in fact, positioned the kan R gene in proximity (within 10 Kbp) of T. pallidum dnaA, dnaN (tp0002), and gyrA (tp0005) genes that, in prokaryotes, are markers for the chromosomal origin of replication (oriC) [20,21]. The >1 ratio kan R :tp0574 in tprA ko -SS14 grown in culture likely reflects partial replication of some chromosomes during propagation, and not differences in amplification efficiency. Such conclusion is also supported by the evidence that in the wild type strain the average tprA:tp0574 ratio, obtained using the same samples above, is 1.31. Overall, these results supported that a) the kan R gene is stably integrated into T. pallidum genome, that b) there are no residual copies of the kan R gene present in episomes, that c) no transformation plasmid is still present and that, for future experiments, d) the copy number of a transgene needs to be compared to that of a neighboring gene to account for replication-induced bias, particularly if the transgene is close to oriC.
In the knockout strain the kan R gene and the tp0574 gene are transcribed by the same promoter. Therefore, cDNA was used to quantify the message level for these two genes. The kan R : tp0574 message ratio was found to be 0.77, which showed the tp0547 gene being slightly more highly expressed than the kan R gene. This result suggested that the choice of using the tp0574 promoter to drive expression of the kan R gene led to very similar message levels for these genes, as hypothesized during the experimental design of the ptprAarms-tp0574prom-kan R plasmid ( Fig 6D). Amplification of kan R message was not detected in the wt-SS14 strain, and no tprA message amplification occurred when cDNA from the knockout strain was used as the template. tprA message was however detected by ddPCR in the wt-SS14 strain. In this case the tprA:tp0547 ratio was 0.040, confirming that the level of transcription of the tprA pseudogene is extremely low, compared to tp0547.

Rabbit infection
We next examined how the tprA ko -SS14 strain acted during in vivo infection in the New Zealand white rabbit model, to determine whether the strain would survive kanamycin treatment

PLOS PATHOGENS
First successful genetic manipulation of T. pallidum of the animal. Only the left testicle was injected with 10 7 treponemes total to facilitate comparison with the healthy right testicle to assess orchitis development. Treated rabbits received the first subcutaneous dose of kanamycin sulfate (5.0 mg/Kg) one hour before infection and every 12 hours thereafter for a total of 10 days. This regimen was hypothesized to be therapeutic based on Veterinary Services recommendation. The rabbit infected with the tprA ko -SS14 strain developed orchitis of the left testicle on day 17 post-inoculation. Treponemal yield from the animal was 1.2x10 8 T. pallidum cells/ml of testicular extract. At the time of harvest, the animal was seropositive with the Treponema pallidum particle agglutination test (TPPA) and the Venereal Disease Research Laboratory (VDRL) test, confirming the establishment of infection. The control rabbit, infected with the wt-SS14 strain but not treated, developed orchitis at day 24 post-infection, and the treponemal yield was 2.2x10 7 T. pallidum cells/ml of testicular extract. This animal was also TPPA-positive but VDRL-negative. On day 24 post-infection, the control rabbit infected with the wt-SS14 strain and subcutaneously treated with kanamycin had not developed orchitis and was euthanized (repeat intramuscular injection of kanamycin was not allowed by local IACUC). Upon analysis of the testicular exudate from this animal, treponemes could be seen, suggesting that the subcutaneous treatment with kanamycin was not completely effective in vivo as it was in vitro (Fig 1B). From this animal, the treponemal yield was however much lower compared to the other rabbits (1.5x10 5 T. pallidum cells/ml, based on detecting 3 treponemal cells in 20 DFM fields). As a further confirmation of treatment failure, this animal was also TPPA-positive but VDRL negative. These data suggested that, although the kanamycin concentration achieved in the rabbits via subcutaneous injection was not completely treponemicidal, tprA ko -SS14 were less susceptible to the antibiotic and proliferated faster than the wild-type. DNA extracted from these treponemal harvests was used for ddPCR targeting the dnaA, tprA, and kan R genes. Droplet Digital PCR showed that for the tprA ko -SS14 propagated in vivo, the kan R :dnaA and tprA:dnaA ratios were 1.05 and 0.007, respectively (Fig 7). For the wt-SS14 strain extracted from the untreated and treated rabbits, respectively, the kan R :dnaA and tprA:dnaA ratios were 0.00 and 1.02, respectively (untreated rabbit) and 0.00 and 0.83, respectively (ineffectively treated rabbit; Fig 7). In these samples the kan R :tp0574 ratio was also obtained and was found to be 1.30 (tprA ko -SS14), and 0.000 (wt-SS14), in both the untreated and ineffectively treated animals. All these in vivo ddPCR results are consistent with ratios seen during in vitro propagation (Figs 3 and 6).

Mass spectrometry
To further confirm the expression of the 31.01 kDa Kan R protein, we performed liquid MS on proteins of the tprA ko -SS14 and wt-SS14 strains separated by SDS-PAGE and ranging approximately from 20-45 kDa. This portion of the T. pallidum proteome was retrieved through of band excision from the acrylamide gel to then undergo in-gel tryptic digestion before MS, and the size range (20-45 kDa) was decided since a distinct~31 kDa band corresponding to Kan R could not be unambiguously identified in the Coomassie-stained gel. Nonetheless, MS data analysis showed that peptides mapping to 77% of the Kan R protein could be isolated from the tprA ko -SS14 strain sample (Fig 8), but not from the paired sample from the wild-type strain. Based on the MS, the amount of the Kan R protein corresponded tõ 1% of the total protein content detected in the specimen. The list of all peptides mapping to the Kan R protein is reported in Table 2. The full MS results for the samples corresponding to the tprA ko -SS14 and wt-SS14 are provided in S1 and S2 Tables, labeled SS14_TprA_KO, and SS14_WT, respectively. In S1 Table, Kan R peptides are labeled PMC_FU_2039. These data demonstrated the expression of the Kan R protein in the tprA ko -SS14 following integration of the gene. No peptides corresponding to the TprA protein (translated using both the +1/+2 reading frames to overcome the effect of the frameshift) were found in the tprA ko -or wt-SS14 samples.

Derivation of a tprA ko -SS14 pure isolate by limiting dilution and increased antibiotic pressure
The tprA ko -SS14 inactivates kanamycin by expressing an aminoglycoside N6'-acetyltransferase that that catalyzes the conversion of kanamycin to N6'-acetylkanamycin using acetyl-CoA. We reasoned that if enough tprA ko -SS14 cells were present, they could bring kanamycin concentration below the minimal inhibitory concentration for this antibiotic and allow growth of wild-type cells. To eliminate the possibility that wt-SS14 T. pallidum cells might survive among the tprA ko -SS14 cells, we performed serial dilutions of the tprA ko -SS14 cells and plated them on multiple wells of a 96-well plate containing Sf1Ep cells. Specifically, wells were seeded with 3300 and 330 treponemal cells from the tprA ko -SS14 culture. Additionally, we increased kanamycin concentration to 200 μg/ml, shown already to be ineffective against the tprA ko -SS14 strain (Fig 5A). Although these cultures were passaged every two weeks to allow treponemal growth, exhausted media was exchanged once a week to ensure Sf1Ep cells viability and replenish the kanamycin supply. Lack of wt-SS14 cells was assessed by performing ddPCR targeting the tprA, kan R , and dnaA genes as described in the Material and Methods section. Results (Fig 9) showed that, after two (Fig 9; Panel 1) and four (Fig 9; Panel 2) weeks in culture, no tprA-specific signal could be obtained by ddPCR, supporting complete lack of wild-type strain in the culture wells, and that in the tprA ko -SS14 strain, the kan R :dnaA copy number was virtually identical.

PLOS PATHOGENS
First successful genetic manipulation of T. pallidum

Discussion
Given the significant impact of syphilis on human health, there is great interest in deepening our understanding of the molecular mechanisms underlying its pathogenesis. In turn, this could lead to improved strategies for disease control and vaccine development. A critical step in this direction is the identification of treponemal factors that contribute to virulence. To date, several experimental approaches have helped syphilis investigators identify and functionally characterize these factors, such as expression in heterologous hosts (e.g. Borrelia burgdorferi, Treponema denticola, and Treponema phagedenis) [22][23][24], comparative genomics [25][26][27][28][29] and, to some extent, gene expression studies and proteomic analysis [16,28,30]. Genetic manipulation strategies, however, particularly those that satisfy Koch's molecular postulates [31], have not been available for T. pallidum. The lack of genetic tools was not overly frustrating because, until 2018 [11][12][13], T. pallidum could not even be continually propagated in vitro. Understandably, this limitation hindered any attempt aimed at genetically engineering this

PLOS PATHOGENS
First successful genetic manipulation of T. pallidum pathogen, as antibiotic selection could not readily be performed. In contrast, some Chlamydia species have been propagated in vitro since the 1950s, and yet genetic tools for this pathogen became available only recently [32]. The protocol we devised to make our vector cross the T. pallidum envelope was indeed inspired by an early protocol used to introduce DNA into Chlamydia [25,32]. The use of CaCl 2 was preferred to that of physical methods such as electroporation simply because the number of treponemal cells yielded by in vitro culture is still very limited compared to other bacteria, including other spirochetes, that have a generation time significantly shorter than T. pallidum and that can be propagated in vitro in axenic cultures. In the SS14 strain, the tprA gene is non-functional due to a frame-shift mutation caused by a CT dinucleotide deletion at position 712 of the annotated gene open reading frame (ORF). The choice to derive a tprA knockout (tprA ko -SS14) strain was driven by the high likelihood that a) this region would not affect T. pallidum viability if removed, being already non-functional in the wild-type SS14 (wt-SS14) strain, and that b) eliminating this pseudogene would not result in a polar effect inhibiting transcription of downstream genes. The tprA locus, even when encoding a functional gene, is likely to be transcribed as a monocistronic mRNA based on prediction software such as Operon-mapper (https://biocomputo.ibt.unam.mx/operon_ mapper/). Ving onto functional genes, the inactivation of many genes might be problematic in the syphilis spirochete, because the small size of treponemal genomes suggests that the genes that escaped evolutionary genomic reduction might be essential. In this scenario, inducible systems for gene silencing or ablation would be helpful. Today, site-directed recombination technologies are increasingly used to manipulate an organism's DNA under controlled conditions in vivo. One example is the system analogous to Cre-Lox recombination but involves the recombination of sequences between short flippase recognition target (FRT) sites by the recombinase flippase (Flp). In addition, an inducible expression system would be beneficial for epigenetic gene silencing approaches, such as the CRISPR/Cas9 system and TALENS [33][34][35][36].

PLOS PATHOGENS
First successful genetic manipulation of T. pallidum Future directions also include the development of shuttle plasmids that do not need to integrate into the T. pallidum genome but are suitable to express T. pallidum ORFs for complementation purposes, for example, but also to express fluorescent reporter proteins such as GFP, CFP, mCherry, or reporter enzymes such as β-galactosidase, with the overall goal of better understanding gene regulation. Most of these goals have already be achieved for other spirochetes and could be potentially adapted to T. pallidum [22,24,[37][38][39]. Expression of luciferase enzymes for in vivo imaging, as done for other spirochetes [23,40], could also be desirable. The transformation of T. pallidum strains with shuttle plasmids should not be complicated by the presence of native plasmids like it is in some Chlamydia species. Plasmids carrying the same origin of replication are generally incompatible and tend not to coexist in a cell. Therefore, the maintenance of a native plasmid limits the introduction of exogenous ones [23,41]. The absence of endogenous plasmids in T. pallidum should eliminate the issue of competition for plasmid replication factors that could hinder the transformation efficiencies of exogenous recombinant plasmids.
The result that the ratio between the kan R gene and tp0574 was higher than originally expected was intriguing and worth discussing in the context perhaps of molecular detection assays for T. pallidum. Following the sequencing of the Nichols strain genome, T. pallidum oriC was localized at nucleotide +1 [42] based on gene synteny. Replication of the bacterial chromosome is initiated at a single oriC region and proceeds in both directions. During growth, replication is generally initiated once per cell cycle, however, under optimal nutrient and media conditions, another round of replication can be initiated even before the previous round has completed, resulting in the inheritance by daughter cells of partially replicated chromosomes [21,43,44]. Partial replication of a chromosome can create an unbalanced ratio between genes that are near oriC (such as kan R , in our case) that are already duplicated, and genes farther away from oriC that are still in single copy (such as tp0574, located at the polar opposite of the kan R gene in the~1Mb T. pallidum chromosome, and therefore one of the last genes to replicate). Hence, during active cell growth, the ratio between copies of genes close to oriC and more distant ones can be remarkably above 1. As mentioned, the >1 ratio kan R : tp0574 in tprA ko -SS14 grown in culture likely reflects partial replication of some chromosomes during propagation. This could support that genes located near the origin of replication could be better targets for T. pallidum detection in clinical samples, but perhaps that genes located further away from the origin could be more suitable to estimate the actual number of treponemal cells in a sample.
Ongoing work in the laboratory is focusing also on analyzing the transcriptional and proteome profile of the tprA ko -SS14 strain compared to the wild-type. It is intriguing that, based on our in vitro cultivation results, the tprA ko -SS14 strain grew significantly faster than the wildtype strain. This difference was not due to errors in the initial inoculum, as the experiment reported in Fig 3 was repeated twice independently, and the results were similar. The ability of the transformed strain to proliferate faster was also suggested by the fact that the rabbit infected with the tprA ko -SS14 strain developed orchitis earlier than the untreated control, even though the inoculum was the same. However, the in vivo experiments are not conclusive in this regard, as only one rabbit was used as untreated control, and the later development of orchitis could simply be due to rabbit-to-rabbit variability. It is nonetheless possible that ablating the tprA pseudogene might have conferred a selective advantage and have reduced the metabolic burden of expressing a gene that is not functional. Previous studies carried on in our laboratory with T. pallidum strains carrying a frame-shifted tprA gene (although not with the SS14 strain), showed that tprA is transcribed in these strains, although at a very low level [17].
Additional experiments to be performed include repeating the transformation experiment reported here, as well as targeting other T. pallidum genes that, unlike tprA, will provide us

PLOS PATHOGENS
First successful genetic manipulation of T. pallidum with the ability to study a phenotype. As proteins mediating motility, antigenic variation, and adhesions are main virulence factors of spirochetes, our attention will focus next on ablating expression of the endoflagella and evaluate whether a non-motile T. pallidum can successfully establish an infection in the rabbit host, on eliminating the tprK donor sites to impair antigenic variation and immune evasion, and knock-out adhesins.

Conclusions
We demonstrate that genetic engineering of the syphilis spirochete is possible with a relatively simple method that has the potential to "transform" our way to approach the study of T. pallidum biology and syphilis pathogenesis.

Ethics statement
Only male NZW rabbits (Oryctolagus cuniculus) ranging from 3.5-4.5 kg in weight were used in this study. Specific pathogen-free (SPF; Pasteurella multocida, and Treponema paraluiscuniculi) animals were purchased from Western Oregon Rabbit Company (Philomath, OR) and housed at the University of Washington (UW) Animal Research and Care Facility (ARCF). Care was provided in accordance with the procedures described in the Guide for the Care and Use of Laboratory Animals [45] under protocols approved by the UW Institutional Animal Care and Use Committee (IACUC; Protocol # 4243-01, PI: Lorenzo Giacani). Upon arrival and before use, all rabbits were bled and tested with a treponemal test (TPPA; Fujirebio, Tokyo, Japan) and a non-treponemal test (VDRL; Becton Dickinson, Franklin Lakes, NJ) to confirm lack of immunity due to infection with Treponema paraluiscuniculi, given that animals are tested randomly by the provider. Both tests were performed according to the manufacturer's instructions. Only seronegative rabbits were used for experimental infection with transformed and wild-type treponemes (see paragraph below for rabbit infection, treatment and sample collection).

Plasmid construct
The pUC57 vector (2,710 bp; Genscript, Piscataway, NJ) was engineered to carry the kan R gene downstream of the T. pallidum tp0574 gene promoter and ribosomal binding site. Appropriate spacing (8 nt) was ensured between the RBS and the kan R gene start codon in the construct. Upstream and downstream of the tp0574prom-kan R hybrid sequence, respectively, two homology arms corresponding to the regions flanking the tprA gene were inserted. The upstream arm was 998 bp in length and corresponded to position 7,343-8,340 of the wt-SS14 strain genome (NC021508.1/ CP004011.1). The downstream arm was 999 bp, and encompassed position 10,165-11,163 of the SS14 genome. The construct was cloned between the XheI and BamHI sites of the pUC57 vector, in opposite orientation compared to the lac promoter that is upstream of the polylinker. Prior to use, the insert underwent Sanger sequencing to ensure sequence accuracy. The sequence of the insert is provided in the S1 Text file. This construct was named ptprAarms-tp0574prom-kan R . Primers annealing a) to the vector only, b) within the cloned insert, and c) upstream of the tprA homology arms in the T. pallidum genome are reported in Table 1. The pUC57 vector carries an ampicillin resistance gene (bla) for selection in E. coli. Because penicillin is the first-line antibiotic to cure syphilis, we first evaluated whether the sequences flanking the bla gene of the pUC57 vector could have had sufficient homology to T. pallidum DNA to induce recombination and integration of the bla gene in the genome, but no homology was found upon BLAST analysis of these regions against the

PLOS PATHOGENS
First successful genetic manipulation of T. pallidum SS14 or other syphilis strain genomes. Regarding the insertion of a kan R gene in the T. pallidum genome, the CDC STI treatment guidelines do not recommend the use of kanamycin for syphilis therapy, hence a kanamycin-resistant syphilis strain does not pose a risk in case of unlikely exposure. To obtain a highly concentrated, endotoxin-free plasmid preparation, the ptprAarms-tp0574prom-kan R was transformed into TOP10 E. coli cells (Thermo Fisher, Waltham, MA), which were then grown first in a 5-ml starter culture overnight, and then in 500 ml of LB media supplemented with 100 μg/ml of ampicillin at 37˚C. The plasmid was purified using the Endo-Free Plasmid Mega Kit (Qiagen, Germantown, MD) according to the manufacturer's instructions. Following purification, plasmid concentration was assessed using an ND-1000 spectrophotometer (Nanodrop Technologies, Wilmington, NC). The vector was then divided into 50 μl aliquots and stored at -80 until use.

Source of T. pallidum for in vitro cultivation, transformation, and selection
The SS14 strain of T. pallidum used for in vitro propagation was obtained from a frozen stock previously propagated IT in NZW rabbits as already reported [46]. This strain was originally isolated in 1977 in Atlanta (USA) from a penicillin-allergic patient with secondary syphilis who did not respond to therapy with macrolides. The SS14 strain was selected because its genetic background is closer to that of contemporary strains that are members of a globally dominant cluster, known as the SS14 clade [27]. In vitro culturing was performed according to Edmondson et al. [11] in the wells of a 24-well plate initially, and then expanded in a 6-well culture plate (Corning Inc, Corning, NY). The microaerophilic atmosphere (MA; 1.5% O 2 , 3.5% CO 2, and 95% N 2 ) necessary to sustain treponemal viability was achieved using a Heracell VIOS 160i tri-gas incubator (Thermo Fisher). Before the addition of the treponemal cells, Sf1Ep cells were incubated in a 5% CO 2 atmosphere in a HeraCell 150 incubator (Thermo Fisher). For transformation, treponemes were first sub-cultured into the wells of a 24-well plate as per protocol [11][12][13]. Briefly, the day before treponemal inoculation, a 24-well plate was seeded with 2x10 4 rabbit Sf1Ep cells/well in 2.5 ml of culture media. The plates were then incubated overnight in the HeraCell incubator. On the same day, TpCM-2 media was prepared according to protocol and equilibrated overnight at 34˚C in the MA incubator. The following day, cell culture media was removed from the 24-well plate, and cells were rinsed with equilibrated TpCM-2 media. Subsequently, each well was filled with 2.5 ml of equilibrated TpCM-2 media, and the plate was transferred to the MA incubator. To prepare the treponemal inoculum, the Sf1Ep cells seeded the previous week with wt-SS14 cells were trypsinized to allow the release and enumeration of spirochetes using the DFM. For treponemal counting, microscope calibration was done according to Lukehart and Marra [47]. A total of 2-3x10 8 treponemes were inoculated 24 hours after plating the Sf1Ep. Following treponemal addition, the total volume of media in each well was brought to 2.5 ml. Two days following treponemal cell addition, the plate was removed from the MA incubator, and 1 ml of old media was replaced with fresh one. Four days after treponemal cell inoculation, the plate was removed again from the MA incubator and the culture media was eliminated gently not to disturb Sf1Ep cells and adherent treponemes and replaced by 500 μl of transformation buffer (50 mM CaCl 2 , 10 mM Tris pH 7.4; equilibrated in MA) containing 15 ug total of ptprAarms-47p-kan R . As a control, to rule out CaCl 2 toxicity to treponema cells, treponemes were also incubated with transformation buffer without plasmid vector. Cells were incubated in these transformation buffers (with and without plasmid) for 10 min at 34˚C in the MA incubator and then washed twice with equilibrated TpCM-2 media to remove free plasmid from the culture wells. Finally, 2.5 ml of fresh TpCM-2 equilibrated in MA were added to the wells, and plates were returned to the MA incubator. The following day, concentrated tissue-culture grade liquid kanamycin sulfate (Sigma-Aldrich, St. Louis, MO) was added to the appropriate wells to reach a final concentration of 25 μg/ml. As a control, to confirm the treponemicidal activity of kanamycin, wild-type treponemes were also incubated in fresh TpCM-2 media containing 25 μg/ml of kanamycin sulfate. Kanamycin sulfate-containing TpCM-2 media was exchanged weekly but treponemes were sub-cultured every two weeks as per published protocol [11] until they reached a density of 3x10 7 cells/ml, counted using the DFM, at which point they were sub-cultured first into one well of a 6-well plate (seeded with 10 5 Sf1Ep cells on the previous day) to upscale the culture, and then into all the wells of a 6-well plate at the following passage to further expand the strain and minimize the chances of culture loss due to contamination. Whenever possible, treponemes that were not used for inoculation of a new plate were pelleted by centrifugation at 15,000 rpm for 10 min using a tabletop centrifuge and resuspended in 1X DNA lysis buffer (10 mM Tris-HCl, 0.1 M EDTA, 0.5% SDS), Trizol (Thermo Fisher), or used to make glycerol stocks, regardless of the number of treponemes counted by DFM.
To compare susceptibility to kanamycin of the tprA ko -SS14 and wild-type strains, treponemes obtained from the tenth in vitro passage were sub-cultured into two 96-well cell culture plates (Corning) instead of 6-well plates to allow a total of eight replicates for each kanamycin sulfate concentration tested. Briefly, the day before inoculation, two 96-well plates were seeded with 3x10 3 rabbit Sf1Ep cells per well in 150 μl of culture media. The plates were then incubated overnight in the HeraCell incubator. On the same day, TpCM-2 media was prepared according to protocol and equilibrated overnight at 34˚C in the MA incubator. On the following day, cell culture media was removed from the 96-well plates, and cells were rinsed with equilibrated TpCM-2 media. Subsequently, each well was filled with 150 μl of equilibrated TpCM-2 media, and the plate was transferred to the MA incubator. To prepare the treponemal inoculum for the 96-well plates, the Sf1Ep cells seeded the previous week with either the tprA ko -SS14 and wild-type strains were trypsinized to allow the release and enumeration of spirochetes. Treponemes were counted using the DFM and diluted in TpCM-2 to 3.3x10 5 T. pallidum cells/ml to obtain a treponemal inoculum of 5x10 4 cells in a total of 150 μl, which were then added to each well of the 96-well plates. The kanamycin sulfate concentrations tested were a 1:2 dilution series ranging from 200 to 1.6 μg/ml, for a total of 8 different concentrations tested in eight replicate wells. No-antibiotic wells, as well as solvent-only wells (water), were also included as controls. Tissue-culture grade kanamycin sulfate was purchased from Sigma-Aldrich. Wild-type and tprA ko -SS14 treponemes in no-antibiotic wells were harvested at day 0 (inoculum), day 1, day 4, and day 7 post-inoculation after incubation at 34˚C in the MA incubator to assess treponemal growth in normal conditions and that, even in absence of antibiotic pressure, the kan R gene would remain steadily integrated in place of the tprA pseudogene in the tprA ko -SS14 treponemes. The experiment where tprA ko -SS14 treponemes were grown in a 96-well plate in the presence or absence of kanamycin sulfate was performed twice to ensure a) the reproducibility, and b) to obtain DNA-free RNA to compare the level of transcription of the kan R gene and that of the tp574 genes by RT-ddPCR.
Treponemes harvested at passage #15 (Week 20 post-transformation) were resuspended in SDS-PAGE sample buffer and proteins were separated on a 12% pre-made acrylamide gel (Thermo Fisher) to evaluate the expression of the Kan R protein using mass spectrometry after gel band excision and digestion (see protocol below).

Rabbit infection, treatment, and sample collection
Pharmaceutical grade kanamycin sulfate (50 mg/ml in water) was prepared by Kelley-Ross compounding pharmacy in Seattle, WA, and stored at -20˚C until use according to the pharmacist's instructions. Once thawed, the bottle was kept at 4˚C and removed from the fridge

PLOS PATHOGENS
First successful genetic manipulation of T. pallidum only to withdraw doses. On the day of infection, tprA ko -SS14 and wild-type strains were harvested from the wells of a 6-well culture plate, enumerated using a Nikon NiU darkfield microscope (Nikon, Melville, NY), and diluted in sterile saline to 3x10 7 /ml. Three NZW rabbits were infected. One rabbit was infected IT only in its left testicle with the tprA ko -SS14 strain. This rabbit received the first subcutaneous dose of kanamycin sulfate (5.0 mg/Kg) one hour before infection and was treated every 12 hours for a total of 10 days, each time with the same dose. The second rabbit was infected with the wt-SS14 strain and also treated with kanamycin as above, while the third rabbit was infected with wt-SS14 but not treated. When orchitis developed, rabbits were euthanized, and the left testicle was removed and minced in sterile saline to extract treponemes. The control rabbit infected with the wt-SS14 strain and treated with kanamycin was euthanized when the other control rabbits developed orchitis, and testicular extracts were also processed. Treponemal suspensions were spun at 1,000 rpm at 4˚C in an Eppendorf 5430R refrigerated centrifuge, and cellular debris were discarded. Treponemes in the supernate were enumerated by DFM and resuspended in 1X DNA lysis buffer by mixing equal volume of extract and 2X buffer. The remaining treponemes were frozen in glycerol stocks (50% serum-saline + 50% sterile glycerol). Terminal bleeding through cardiac puncture was performed to obtain serum for VDRL and TPPA tests, performed as described above. All sera were heat-inactivated at 56˚C for 30 min before use. Kanamycin dosage was selected based on Veterinary Services recommendation on the use of this antibiotic to rid of organisms such as E. coli, Salmonella spp., Klebsiella spp., and Pasteurella spp. in mid-size animals such as cats. The frequency of treatment was decided based on data reported in [48].

DNA and RNA extraction
DNA extraction from cultured tprA ko -SS14 or wild-type strains propagated in 6-well plates following the transformation procedure was performed using the QIAamp mini kit (Qiagen) according to the manufacturer's instructions. Extracted DNA was stored at -80˚C until use for qualitative PCR, ddPCR, or WGS (see below). DNA extraction from the 96-well culture plates used to assess susceptibility to kanamycin of the tprA ko -SS14 and wild-type strain, respectively, was performed after a one-week incubation of the cells in the MA incubator. Plate culture media was removed from each well with a laboratory aspirator connected to the vacuum line using a 0.22 μm filter and discarded. Cells were not trypsinized but 200 μl of Genomic Lysis Buffer (Zymo Research, Irvine, CA) for DNA extraction was added. Cells were then lysed through incubation in lysis buffer for 30 min at room temperature as per provided protocol. While propagating the SS14 strain, we used dark-field enumeration of treponemes present in the cell culture supernate and attached to Sf1Ep cells but released by cell trypsinization, and determined that the majority (~85%) of T. pallidum cells in vitro adhere to the rabbit epithelial cell monolayer, similar to what was reported by Edmondson et al. for other cultivated strains [11]. This evidence allowed us to discard the culture media without concern that the experimental results would be significantly affected. Following cell lysis, the plates were frozen at -20˚C until extraction could be completed. To purify DNA, the 96-well plates were thawed at 56˚C in a dry incubator and quickly spun to recover condensation drops on the well lids. DNA was extracted using a Quick-DNA 96 kit (Zymo Research) according to the manufacturer's protocol. DNA was eluted in 100 μl of water and stored at -20˚C until amplification using qPCR to evaluate treponemal load. The same samples were used to perform ddPCR (see protocol below) targeting tp0001 (dnaA), tprA, and the kan R genes to investigate the ratio between these targets. As an additional control, extracted samples were also tested randomly using pUC57-specific primers (Table 1) annealing upstream and downstream of the cloned insert to assess carry-over of the plasmid used for transformation.
RNA extraction was performed using the Quick-RNA 96 kit (Zymo Research) following the manufacturer's instructions with the exception that in-column DNA digestion using the DNaseI enzyme (provided by the kit) was prolonged for a total of 1 hour and performed using 50% more enzyme than normally suggested. Single samples from in vivo and in vitro propagation were instead extracted according to the Trizol reagent manual. Total RNA was treated with DNaseI according to the protocol provided with the TURBO DNA-free kit (Thermo Fisher). DNA-free RNA was checked for residual DNA contamination by qualitative amplification using primers specific for the tp0574 gene (primers in Table 1) as already described [49]. Reverse transcription (RT) of total RNA was performed using the High-Capacity cDNA Reverse Transcription kit (Thermo Fisher) with random hexamers according to the provided protocol. cDNA samples were stored at -80˚C until use for qualitative PCR to demonstrate expression of the kan R gene or for ddPCR to quantify the level of expression of the tp0574 and the kan R genes (primers in Table 1).

Qualitative and quantitative PCR
Samples harvested during routine propagation of the tprA ko -SS14 and wild-type strains were assessed for integration of the kan R gene into the tprA locus by using qualitative PCR. In the first amplification, the sense primer targeted a region of the T. pallidum genome immediately upstream of the left tprA homology arm of the vector (and hence not cloned into ptprAarms-tp0574prom-kan R ), while the antisense primers targeted the kan R gene, with the rationale that only a kan R gene integrated into the tprA locus would provide amplification. Primers and amplicon size are reported in Table 1 and schematically represented in Fig 2, along with the results of the PCR done on DNA extracted from treponemes at Passage #7. In the second PCR, the sense primer targeted the kan R gene, and the antisense primer (Table 1) targeted the genomic region downstream of the left tprA homology arm of the vector. Amplification of the tp0574 gene was used as positive amplification control. Amplifications were performed using five microliters of extracted DNA in 50 μl final volume containing 2.5 units of GoTaq polymerase (Promega, Madison, WI), 200 μM of each dNTP, 1.5 mM of MgCl 2 , and 400 nM of sense and antisense primers. Cycling parameters were initial denaturation (94˚C) and final extension (72˚C) for 10 min each. Denaturation (94˚C) and annealing (60˚C) steps were carried on for 1 min each, while the extension step (72˚C) was carried out for 1 or 2 min depending on amplicon length. A total of 40 cycles were performed in each amplification. A qPCR assay was instead used to quantify treponemal load in samples extracted from the 96-well plate following the kanamycin susceptibility assay of the tprA ko -SS14 and wild-type strains. In this case, the treponemal load was evaluated using a qPCR approach targeting the tp0574 gene previously described [17]. Primers are reported in Table 1. Briefly, an absolute quantification protocol using an external standard was used to quantify the tp0574 gene copy number at the time of sample harvest. Standard construction was also previously described in detail [17]. For amplification, the Powerup SYBR Green Master Mix (Thermo Fisher) was used. Amplifications were run on a QuantStudio 5 thermal cycler (Thermo Fisher) and results were analyzed using the instrument software. Data were imported into Prism 8 (GraphPad Software, San Diego, CA) and further analyzed to assess statistical significance of the values from test and no-antibiotic control groups using one-way ANOVA with the Dunnett test for correction of multiple comparisons or t-test, with significance set at p<0.05 in both cases.

Droplet digital PCR (ddPCR)
Droplet digital PCR assays were conducted to assess the ratio between the number of copies of the kan R gene and another target on T. pallidum genome in the tprA ko -SS14 strain, in this case, the tp0574 gene, the tp0001 gene (dnaA), or the tprA gene. The tp0574 and dnaA targets were chosen because of their relative distance to the kan R insertion site (< 10 Kbp from dnaA, and~513Kbp from the tp0574 gene) to account for possible discrepancies due to the vicinity of the kan R gene to T. pallidum chromosomal origin of replication. ddPCR was also used to evaluate the level of transcription of the kan R gene and the tp0574 gene, as both genes are transcribed by the same promoter. Amplification of the tprA gene was used as control. To this end, four sets of primers/probes (Table1) were designed. DNA obtained from the 96-well plate used to perform the kanamycin susceptibility assay with the tprA ko -SS14 strain was used along with the DNA from the tprA ko -SS14 strain obtained from Passage #8, and DNA extracted from the tprA ko -SS14 and SS14 wild-type treponemes propagated in vivo. cDNA was obtained (as described above) from the 96-well plate used to perform the kanamycin susceptibility assay with the tprA ko -SS14 strain, and from the tprA ko -SS14 and SS14 wild-type treponemes propagated in vivo and in vitro (Passage #8).
ddPCR was performed on a Bio-Rad QX100 system (Bio-Rad, Carlsbad, CA). Each reaction was performed using ddPCR Supermix for Probes (Bio-Rad) with the final concentration of primers at 900 nM and probes at 250 nM in a total reaction volume of 25 μl. Before amplification, template DNA was digested with 25 units of EcoRI (New England Biolabs, Ipswich, MA). cDNA was used without the digestion step. After droplet generation, droplets were transferred to a 96-well PCR plate and amplified on a 2720 Thermal Cycler (Thermo Fisher) with the following cycling parameters: 94˚C for 10 min, followed by 40 cycles of 94˚C for 30 s and 60˚C for 1 min, and 98˚C hold for 10 min. After amplification, the plate was transferred to QX200 droplet reader (Bio-Rad). Results were analyzed using the QuantaSoft software (Bio-Rad).

Analysis of Kan R expression by mass spectrometry
Expression of the KanR protein (molecular weight of 31.1 KDa) was assessed by liquid chromatography-mass spectrometry (LC-MS) in tprA ko -SS14 and wild-type treponemes. For the SDS PAGE, the tprA ko -SS14 and wild-type treponemes were harvested as described above from culture plates (Passage #15, Week 20 post-transformation), counted by DFM, and centrifuged at 100xg (1,000 rpm in a tabletop centrifuge with a 9 cm radius) to remove residual Sf1Ep cells. The resulting supernate was then spun at 15,000 rpm at RT for 10 min to pellet treponemes and pellet resuspend in 1X SDS-PAGE sample buffer (50 mM Tris-HCl; 100 mM DTT; 70 mM SDS; 1.5 mM Bromophenol blue, 2M glycerol). Approximately 10 9 treponemes were resuspended in a final volume of 200 μl of SDS-PAGE sample buffer. Samples were boiled and loaded onto a 12% precast Tris-Tricine gel in a mini-Protean apparatus (both from Bio-Rad, Hercules, CA). Gels were stained using SimplyBlue SafeStain (Thermo Fisher). Subsequently, a gel segment encompassing protein sizes between 20-40 kDa (and thus including the~30 kDa Kan R protein) was excised and bands were subjected to overnight in-gel trypsin digestion as previously described [17,27]. The volume of digestion products was reduced to approximately 10 μl using a speed-vac. Peptides were analyzed by LC-MS at the Fred Hutchinson Cancer Research Center proteomics facility using an LTQ HP1100 mass spectrometer (Thermo Fisher), results were analyzed using the Proteome Discoverer software. Identified peptides were filtered to a false discovery rate (FDR) of <0.01 to ensure high confidence in the identified peptides.

Whole-genome sequencing
Whole-genome sequencing was performed following DNA extraction of tprA ko -SS14 and wild-type treponemes cultured in vitro (Passage #9; Fig 1B). Pre-capture libraries were prepared from up to 100 ng input DNA using the KAPA Hyperplus kit (Roche) and TruSeq adapters and barcoded primers (Illumina), following the manufacturer's protocols, yielding an average fragment size longer than 500 bp. Hybrid capture of T. pallidum genomic DNA was performed overnight (>16 hours) using a custom IDT xGen panel designed against the reference genome NC_ 010741, following the manufacturer's protocol. Short-read sequencing was performed on an Illumina MiSeq with 192 bp single-end reads, yielding at least 1e6 reads per sample. Reads were adapter and quality trimmed using Trimmomatic v0.39 [50] and assembled to the TPA reference genome NC_021508.1, or NC_021508.1 with the tprA gene replaced by the kanamycin cassette, using Bowtie2 [51], to an average coverage exceeding 175x across the genome for all samples. Manual confirmation of expected coverage and junctions was performed by visual inspection in Geneious Prime v2020.1.2 [52].
Derivation of a tprA ko -SS14 pure isolate by limiting dilution and increased antibiotic selective pressure Sf1Ep cells were seeded into the wells of a 96-well culture plate at a density of 3,000 cells/well in 150 μl of MEM and cultured overnight in a 37˚C in the 5% CO 2 incubator. The next morning after the cells had adhered to the plate the MEM was exchanged for 135 μl TpCM2 media which had been equilibrated overnight in a 34˚C in the MA incubator. The cells were equilibrated in the MA incubator for at least 3 hr before proceeding to inoculation of the tprA ko -SS14 cells, harvested using trypsinization from a 6-well plate where the strain was routinely propagated. Following harvest, cell concentration was determined by DFM. For the serial dilutions, 3.3x10 4 tprA ko -SS14 cells were inoculated into each of 8 wells of the prepared 96-well plate, mixed thoroughly, and then 15 μl of the diluted T. pallidum mixture moved to the next set of 8 wells. In this manner, successive 1:10 dilutions of tprA ko -SS14 were made on the plate until the mixture was diluted to 0.33 tprA ko -SS14 cells/well. Kanamycin was added to each well for a final concentration of 200ug/ml. The plate was grown for 2 weeks in the MA incubator, with a half-media change at one week. After 2 weeks, the supernatant was removed from the plate, the wells washed briefly with 20ul of trypsin, and then incubated with 20ul of trypsin for 5min at 37˚C to release adherent T. pallidum cells. Ten microliters of the resulting T. pallidum mixture were inoculated into a fresh 96-well plate prepared and incubated as described above. To the remaining 10 μl, 200 μl of genomic lysis buffer from a Quick-DNA 96 kit (Zymo Research, Irvine, CA) was added. The plate was sealed and stored at -20˚C until DNA extraction. DNA was extracted using the Quick-DNA 96 kit (Zymo Research). DNA was eluted in 30 μl of molecular H 2 O, the minimum volume recommended for the kit. Lack of amplification signal for tprA was assessed by ddPCR as described above.
Supporting information S1