In this study, we have compared the growth rates of seven T. pallidum strains under in vitro conditions. As only a limited number of live strains were successfully isolated from patients, propagated in rabbits, and stored as viable cells6,29,30,31, our study mapped a considerable part of the available T. pallidum cultures.
All T. pallidum strains grew continuously in vitro for a period over two years, which suggests that they have been well adapted to in vitro conditions. Despite this fact, we observed differences among in vitro growth of Nichols-like and SS14-like strains. Especially, Philadelphia 1 and Grady grew significantly slower compared to other strains.
The results of standard subcultures (20× dilution every week) in long-term cultivation were highly variable, likely as a result of different numbers of inoculated treponemes during individual subcultures and/or other factors. Indeed, the fold increase of T. pallidum was dependent on the dose of inoculated treponemal cells, although only for faster-growing Nichol-like strains. The absence or minimal source limitation of SS14-like strains during 7-day-long subcultures likely reflects the slower in vitro growth of SS14-like strains. In fact, these strains reached the point of source limitation slightly later, between 8th and 10th day of cultivation (Supplementary Figure S2). Moreover, source limitation of SS14-like strains was observed during enrichment of subcultures (5–10× dilution every week) (see Supplementary Figure S1). Given that treponemal cells compete with eukaryotic feeder cells for nutrients, the observed differences in T. pallidum saturation densities may reflect the depletion of nutrients caused by the feeder cells.
For minimization of the effect of source limitation in cultivation well, we calculated the minimal generation times for each strain, which should be closer to real generation time than average/median values. The estimated generation time of 32.6–33.5 h for Nichols-like strains and 36.8–39.8 h for SS14-like strains is generally consistent with the previously reported values for strains Nichols, Mexico A, SS14, UW231B, and UW249B16,17,18. Moreover, the calculated in vitro generation times are higher but relatively close to the T. pallidum generation times estimated during experimental rabbit infections (~ 30 h)19,32,33.
Parallel cultivations of monocultures with a defined number of inoculated treponemal cells were performed to obtain more precise quantification of growth differences. Here, the significantly fastest strains DAL-1 and Madras were followed by Haiti B, Mexico A, SS14, and finally Philadelphia 1 and Grady. Identification of three subgroups of strains with different growth rates mostly agrees with the growth rates determined from the long-term cultivation and suggested that the standard subcultures were partially distorted by the presence of source limitation, most prominent among the fast-growing strains.
Despite the more standardized conditions of in vitro cultivation in comparison to animal infection19, the growth of T. pallidum in parallel wells with identical components could potentially result in inconsistencies due to the inherent complexity of growth conditions. For example, Edmondson et al. demonstrated that the growth of T. pallidum is influenced by the number of viable/attached rabbit cells17. Indeed, the results of parallel cultivation of treponemal strains revealed a relatively high extent of growth variability among experimental replicates (Supplementary Table S5). Accordingly, the in vitro co-cultivation scheme was used to avoid subtle differences in the number of rabbit cells, their age and attachment rate, and other factors that could contribute to the observed variability in parallel experiments. Furthermore, we analyzed day 7 (and day 28) of the cocultures to minimize potential differences in the viability of individual strains regarding their resistance to atmospheric oxygen during the preparation of T. pallidum mixtures on day 0.
Using binary co-cultivation of T. pallidum strains, we revealed detectable differences in the growth rates among all tested strains. The DAL-1 strain grew at the fastest rate, followed by Madras, Mexico A, Haiti B, SS14, Grady, and Philadelphia 1. In comparison to the standard continual and parallel monocultures, binary co-cultivation revealed that each strain has its own in vitro growth rate substantially differing among strains. The internal congruency in individual co-cultivations was found for all performed combinations with no evident growth antagonisms among the strains. Concurrently, the quantitative errors of co-cultivations were found higher with the increased distance from the T. pallidum reference comparison indicating a fair agreement of all comparisons. Moreover, the differences in the growth rates were generally consistent with the phylogenetic distance of the analyzed T. pallidum strains suggesting that the individual growth rates reflect genomic differences of each strain.
In general, Nichols-like strains grew faster in vitro compared to SS14-like isolates, with exception of Haiti B and Mexico A. However, both strains are somewhat exceptional. Haiti B was originally considered a T. pallidum ssp. pertenue, the causative agent of yaws, and was shown to be able to grow in hamsters34. Later, genomic analysis showed that Haiti B belongs to the Nichols-like cluster of syphilis-causing strains, but it differs substantially from the Nichols strain. The difference in more than 200 nucleotide positions (CP032623.1 vs. CP004010.2) represents a relatively significant divergence even for the Nichols-like strains. Mexico A harbour two recombinant loci from T. pallidum ssp. pertenue and/or T. pallidum ssp. endemicum35 and can be considered a rather exceptional SS14-like strain. To date, no clinical isolate related to Haiti B (allelic profile 9.83.10) has been found (PubMLST36. Clinical isolates related to Mexico A (allelic profile 1.13.10) were only sporadically found in North or South America9,37.
It is unclear whether the observed differences in growth rates reflect the in vitro growth in the presence of human cells and, more importantly, during infections of humans. However, there is currently no available in vitro cultivation system that uses human cells instead of rabbit cells. A further limitation of this study is that T. pallidum growth in vitro remains suboptimal, with growth rates lower than those observed in vivo in rabbits19,32,33. It is important to note that in vitro cultivation was not completely optimized for individual strains, and binary co-cultivation was performed in only one biological experiment. These facts could therefore affect the determined growth characteristics.
Our findings suggest that the growth of each strain represents an evolutionary selected optimum. Based on several recent epidemiological studies6,9,10,38,39,40,41,42, SS14-like strains are prevalent worldwide (80–90%); especially in European population. Moreover, frequent contemporary isolates around the world (allelic profiles 1.1.1, 1.1.3, 1.1.8, and 1.3.1)9,37,40,41,42,43 are closely related to Philadelphia 1 and Grady (both 1.1.1). These findings suggest that the T. pallidum isolates with slow growth in vitro are prevalent among the infected human population. However, further studies will be needed to discover the molecular basis of the observed differences in the T. pallidum growth rates.
Strains Philadelphia 1 and Grady grew considerably slower than SS14, despite the fact that all three genome sequences are highly related, with just over a dozen nucleotide differences per genome comparison. Besides four single-nucleotide polymorphisms in the intergenic regions8, they differ in six protein-coding genes, including TP0117 (TprC), TP0341 (MurC), TP0488 (Mcp2), TP0620 (TprI), TP0705, and TP0793, which may be involved in the observed growth differences. Nevertheless, other explanations such as occurrence of adaptive mutations in the SS14 strain cannot be excluded at this time.
Our study showed that the in vitro growth rate of T. pallidum strains is an inherent, individual characteristic of each strain, revealing significant growth differences across a comprehensive set of T. pallidum strains. These growth rate distinctions could reflect more profound physiological variations among syphilis-causing strains, potentially contributing to the variability in syphilis symptomatology that has long been observed among patients.