Susceptibility to Penicillin G and Ceftriaxone in 3 Clinical Treponema pallidum Isolates Is Not Altered by Amino Acid Polymorphisms in the Tp0705 Penicillin Binding Protein

Abstract

We demonstrated no differences in susceptibility to penicillin G and ceftriaxone in 3 modern Treponema pallidum isolates (UW244B, UW249B, and UW330B), each carrying a variant of the penicillin-binding protein Tp0705. This suggests that these polymorphisms should not be a reason for concern when β-lactams are prescribed for syphilis treatment.

We report that 3 modern Treponema pallidum isolates carrying amino acid polymorphisms at positions 506, 625, and 708 of the penicillin-binding protein Tp0705 are equally susceptible to penicillin G and ceftriaxone.

Syphilis, caused by the spirochete Treponema pallidum subspecies pallidum (T. pallidum), remains a global health concern, with a rising incidence worldwide^1,2 and significant challenges in disease control. Once a syphilis diagnosis is established, the recommended treatment is penicillin G in different formulations and dosages, depending on disease stage and invasion of the central nervous system.³ In addition to penicillin G, other natural or synthetic β-lactam antibiotics have been shown to be effective against T. pallidum in preclinical⁴ and clinical studies. In observational studies, for example, amoxicillin was highly effective in treating early syphilis.^5,6 Ceftriaxone, a potent third-generation β-lactam cephalosporin, is listed as a possible alternative to penicillin G to treat early syphilis, and data are available to support its efficacy for neurosyphilis.^7–9 Promising results are also available on the efficacy for early syphilis treatment for cefixime, another third-generation cephalosporin,^10,11 although more studies are needed.

The reliance on β-lactams for syphilis treatment is justified by the lack of reliable evidence of resistance to this class of antibiotics after 8 decades of use, consistent with their exceedingly low minimal inhibitory concentration (MIC) when tested in vitro against T. pallidum.^4,12 In contrast, there was rapid development and spread of resistance to macrolides when oral azithromycin was introduced for syphilis treatment.¹³ The T. pallidum genome does not contain plasmids or other mobile elements, which makes the acquisition and retention of environmental DNA coding for bona fide β-lactamases or other antibiotic resistance genes very unlikely. Nonetheless, we should not dismiss the possibility that amino acid polymorphisms on the pathogen penicillin-binding proteins (PBPs) could decrease the efficacy of β-lactams or even induce resistance, as hypothesized by Cha et al.¹⁴ or by Mi et al.¹⁵ These investigators, however, could not demonstrate their hypotheses due to lack of isolates with the proper genotypes to study. Furthermore, in 2025, 2 laboratory-derived mutant T. pallidum strains carrying the A1873G (inducing the amino acid change M625V) mutation in the PBP-encoding tp0705 gene were reported to be partially resistant to ceftriaxone and penicillin G when exposed to low concentrations of these antibiotics,¹⁶ albeit with modest absolute effect sizes. Here, we tested the in vitro susceptibility to penicillin G and ceftriaxone of 3 T. pallidum clinical isolates (UW244B, UW249B, and UW330B) attained from patient blood samples, each carrying a distinct variant of Tp0705, due to polymorphisms at positions 1516, 1873, and 2122 of the gene, corresponding to amino acid positions 506, 625, and 708 of the Tp0705 protein (Table 1) to see whether A1873G/M625V and other naturally occurring polymorphisms on the Tp0705 protein would alter susceptibility to β-lactams.

TABLE 1.

Polymorphisms in the tp0705 Gene/Tp0705 Protein in the Tested Strains

	Nucleotide Position 1516/Amino Acid Position 506	Nucleotide Position 1873/Amino Acid Position 625	Nucleotide Position 2122/Amino Acid Position 708
UW244B	Guanine/alanine	Adenine/methionine	Adenine/serine
UW249B	Guanine/alanine	Guanine/valine	Adenine/serine
UW330B	Adenine/threonine	Adenine/methionine	Guanine/glycine

METHODS

T. pallidum Strains and In Vitro Propagation

The UW244B, UW249B, and UW330B T. pallidum strains were isolated in 2004 (UW244B and UW249B) and 2005 (UW330B) by Dr. Christina Marra at the University of Washington via rabbit intratesticular injection with whole blood from patients with secondary (UW249B), early latent (UW330B), and late latent/unknown duration (UW244B) infection.¹⁷ For this study, all strains were grown in vitro as previously described¹⁸ to perform antibiotic susceptibility testing. Genome sequencing data are available under BioProject/ID PRJEB20795/ERR2756217 (UW244B), PRJNA313497/SRR3268758 (UW249B), and PRJEB20795/ERR2756199 (UW330B) in International Nucleotide Sequence Database Collaboration databases. Polymorphisms were also confirmed by amplification/Sanger sequencing of the tp0705 gene (see Supplementary Methods, http://links.lww.com/OLQ/B317). The UW244B, UW249B, and UW330B isolates are related to SS14-clade strains (n = 1582 sequenced genomes, including those that belong to strains form patient samples collected up to 2020) from each continent by a median of 10 or fewer core genome SNPs and are therefore representative of circulating strains.¹⁷

Penicillin G and Ceftriaxone Susceptibility Testing

Susceptibility testing of the 3 selected strains was performed as previously described⁴ by incubating in vitro for 7 days T. pallidum with increasing concentrations of penicillin G or ceftriaxone above and below the MIC previously determined for these antibiotics.^4,19 Concentrations ranged from 0.125 to 12 ng/mL (penicillin G), or 0.625 to 25 ng/mL (ceftriaxone), followed by determination of treponemal growth by quantitative polymerase chain reaction in test and control (i.e., no antibiotic and antibiotic diluent only) replicate culture wells. In both assays, penicillin G at 60 ng/mL was used as a treponemicidal control. Penicillin G (catalog no. J63032) and ceftriaxone (catalog no. C5793) were purchased from Thermo Fisher (Waltham, MA) and Millipore Sigma (Burlington, MA), respectively. The susceptibility assay is described in more detail in Supplementary Methods, http://links.lww.com/OLQ/B317. A 2-way analysis of variance test was used to calculate statistical significance (*P = 0.01–0.05) among medians within a concentration group.

RESULTS

Incubation of the 3 strains with penicillin G resulted in complete killing of T. pallidum when exposed to a concentration of 1.00 ng/mL or higher of antibiotic (Fig. 1A). Below 1.00 ng/mL, some treponemal growth occurred, although not as efficiently as when the antibiotics were absent from the culture media, consistent with previous reports that identified the secondary MIC for penicillin G (i.e., the lowest antibiotic dilution at which growth occurred but at a significantly lower rate compared with the no-antibiotic control wells) to be 0.5 ng/mL.¹⁹ In one instance (Fig. 1A), we found a significant difference in growth between the UW249B and UW330B strains exposed to 0.25 ng/mL of penicillin G. This difference, however, was not confirmed at any other nontreponemicidal concentration tested. Similarly, exposure to ceftriaxone concentrations equal to or above 2.50 ng/mL was fully bactericidal in vitro (Fig. 1B), which was also consistent with previous studies,⁴ whereas a single significant difference in growth between the UW244B and UW249B strains could be seen upon exposure to 0.625 ng/mL of antibiotic (Fig. 1B), but not confirmed at any other concentration, which was not fully bactericidal.

Figure 1.

Susceptibility to Penicillin G and Ceftriaxone of T. pallidum Strains Carrying Tp0705 Variants.

DISCUSSION

Penicillin-binding proteins involved in cell wall synthesis are the targets of β-lactam antibiotics, and mutations/polymorphisms in PBPs can reduce susceptibility to penicillin or induce resistance by decreasing the affinity for the drug of the transpeptidase domain of these enzymes.²⁰ Although T. pallidum has been treated with penicillin for decades, and penicillin resistance has never been experimentally demonstrated, rare reports of treatment failures following administration of β-lactams exist.^15,16,21,22 Given the pivotal role of β-lactams in syphilis treatment, evaluating whether allelic variants of the 5 known T. pallidum PBPs (the Tp0500, Tp0547, Tp0574, Tp0705, and Tp0760 proteins) confer differential susceptibility to β-lactams is a worthy task, now possible via a novel cultivation system.¹⁸

The isolates analyzed here did not show significant differences in the penicillin G susceptibility assay (Fig. 1A), although UW249B grew slightly less when exposed to 0.250 ng/mL of penicillin G. This difference, however, was not seen at immediately higher (0.5 ng/mL) or lower (0.125 ng/mL) concentrations, suggesting that the antibiotic concentration range at which this phenomenon is detectable might be very limited.

When a patient receives a single intramuscular (IM) injection of 1.2 M units of benzathine penicillin G (half the total dose a patient with early syphilis would receive) and the blood plasma concentration is measured daily over time, the recorded values range from 119 ng/mL (24 hours postinjection, corresponding to ~238 times the penicillin G MIC for T. pallidum) to 5.0 ng/mL (at 29 days postinjection, corresponding to ~10 times the MIC).²³ All concentrations in the 119–5.0 ng/mL range are highly treponemicidal for T. pallidum in vitro, as shown independently by us and other investigators.^4,19 Although no data on the concentration of penicillin G in tissue could be found after IM injection, even if the different Tp0705 alleles studied here could modulate susceptibility to penicillin G, this effect would manifest at antibiotic concentrations that are too low to be clinically relevant, given that one single 2.5 M unit dose of benzathine penicillin G administered intramuscularly is sufficient to treat uncomplicated syphilis.²⁴

A similar conclusion can be drawn for ceftriaxone, for which a plasma concentration of 80.5 μg/mL (~32,200 times the reported ceftriaxone MIC for T. pallidum)⁴ and 29.7 μg/mL (~11,880 times the MIC) are achieved by a single IM dose²⁵ of 1 g at 1 and 24 hours postinjection, respectively. Far higher plasma concentrations were reached by intravenous (IV) injection of ceftriaxone in the same experiment.²⁵ In our study, slight differences in strain growth upon ceftriaxone exposure could be seen only at the lowest concentration tested (0.625 ng/mL; Fig. 1B), although such differences were not confirmed at higher concentrations. Considering that ceftriaxone therapy for syphilis requires a 10-day daily administration regimen, either IM or IV, the difference seen here (Fig. 1B) also seems negligible from a disease management standpoint.

Interestingly, the polymorphism associated with decreased susceptibility of the laboratory-derived strains used in Pospíšilová et al.,¹⁶ A1873G/M625V, did not seem to confer a similar advantage to strain UW249B, also carrying the same polymorphism, in our study. This, in our opinion, reiterates the importance of using clinical isolates in assessing drug susceptibility whenever possible. In the past, we also tested the susceptibility to ceftriaxone of the historical SS14 strain that, like the widely used Nichols strain, carries the A506, M625, and G708 polymorphisms in Tp0705.⁴ These earlier data showed that for this strain, like the ones studied here, a ceftriaxone concentration equal or above 2.5 ng/mL is treponemicidal.

Furthermore, we recently performed an azithromycin susceptibility experiment using the UW330B and SS14 isolates, both known to be macrolide resistant via different mutations (A2058G or A2059G) in the 23S rRNA gene.⁴ These experiments showed that, when exposed to 64 times the azithromycin MIC for T. pallidum, these isolates grew as if no antibiotic was present in the media, further supporting that, although UW330B and most of the modern syphilis strains circulating worldwide are indeed fully resistant to azithromycin,¹³ the same cannot be said for resistance to β-lactams in the isolates tested in this study based on our experimental results.