INTRODUCTION
Syphilis is a multistage sexually transmitted infection of worldwide importance, with estimates of total disease burden ranging from 18 to 56 million individuals (
1–4).
Treponema pallidum subsp.
pallidum (
T. pallidum), the causative agent of syphilis, was first identified by Schaudinn and Hoffman in 1905 (
5,
6) as “very light, thin spiraled microorganisms, turning around their largest length and moving back and forth.” Rapid progress in the study of this bacterium was made within 5 years, with the verification of the presence of spirochetes in experimentally infected animals by Metchnikoff and Roux (
7), the invention of dark-field microscopy for easy visualization of
T. pallidum by Karl Landsteiner, development of the first serological test for syphilis by von Wassermann et al. (
8), and the introduction of arsphenamine as an effective, relatively nontoxic antisyphilis agent by Paul Ehrlich (
9,
10). Successful culture of
T. pallidum was reported almost immediately and during the subsequent decades (
11,
12), but these reports were found to be either irreproducible or the result of contamination with nonpathogenic
Treponema species that colonize human skin (
13). During the 1970s, progress was made in characterizing some
T. pallidum physiological properties, most notably its microaerophilic nature and improved survival in the presence of mammalian cells (
14).
In 1981, Fieldsteel, Cox, and Moeckli (
15) reported the consistent occurrence of up to 100-fold multiplication of
T. pallidum in a coculture system consisting of Sf1Ep cottontail rabbit epithelial cells, a modified tissue culture medium with heat-inactivated fetal bovine serum, dithiothreitol (DTT) as a reducing agent, and a microaerobic atmosphere containing 1.5% O
2. These results were reproduced in hundreds of experiments (primarily by the Cox and Norris groups) and reported in over 25 publications (reviewed in references
14 and
16). However, treponemal multiplication and survival were limited to 12 to 18 days, despite efforts to refine this system. Attempts to subculture
T. pallidum provided little improvement in the cumulative fold increase or survival of the bacterium (
14,
16,
17). The same limitation has existed for the closely related organisms that cause yaws (
T. pallidum subsp.
pertenue), bejel (
T. pallidum subsp.
endemicum), pinta (
Treponema carateum), and venereal spirochetosis in rabbits and hares (
T. paraluiscuniculi) (
18,
19). The inability to culture these organisms continuously
in vitro has necessitated their propagation in rabbits for use in research, greatly hindering investigation of these important pathogens.
In this study, we utilized a modification of the method described by Fieldsteel et al. (
15,
16) to achieve reproducible, long-term multiplication of
T. pallidum subsp.
pallidum in a tissue culture system. In ongoing experiments,
T. pallidum multiplication and full viability have been maintained for up to 27 passages over a period of 6 months. Results obtained with the Nichols reference strain have been verified using two recent
T. pallidum isolates. The
in vitro-cultured treponemes retained their characteristic ultrastructure (as determined by cryoelectron microscopy) and full infectivity (as demonstrated by rabbit inoculation experiments). The availability of this culture system is likely to lead to a better understanding of
T. pallidum physiology, structure, gene expression, regulatory pathways, pathogenesis, immunologic properties, and antimicrobial susceptibility.
DISCUSSION
Here, we report the first consistent long-term in vitro cultivation of T. pallidum subsp. pallidum. The successful culture of recent syphilis isolates UW231B and UW249B as well as the long-established Nichols strain indicates that the cultivation procedure will likely be applicable to other syphilis isolates. Given the extreme similarity of all the T. pallidum subspecies, T. carateum, and T. paraluiscuniculi, it is probable that the same conditions will be effective in propagating all the members of this group of pathogens, as will be examined in future studies.
The culture method used was strongly based on the prior studies by A. Howard Fieldsteel, David L. Cox, and their coworkers, who systematically established the requirements for growth of
T. pallidum in primary cultures. The extended
in vitro survival of
T. pallidum in the presence of mammalian cells had been noted by several groups (
32–37), and examination of several cell cultures indicated that Sf1Ep cells performed better than other cell types in this aspect (most likely because of their low growth rate and low metabolism). Fieldsteel et al. (
37) used oxygen gradients formed in Leighton tubes to demonstrate that microaerobic conditions further prolonged survival and apparent multiplication. It was also noted that only certain commercial fetal bovine serum (FBS) lots supported
T. pallidum viability (
38). By combining these components with Eagle’s MEM and DTT to decrease the presence of reactive oxygen species, consistent
in vitro multiplication was achieved and reproduced in many experiments. However, attempts to obtain continued growth through subculturing were generally unsuccessful, with the combined yields of the primary and passaged cultures rarely exceeding that of the primary culture alone. Cox obtained enhanced survival and growth with subculture in 4 experiments in which up to 2,000-fold increases over a 17-to-30-day period were observed (reviewed in reference
14). However, these results were not reproduced in subsequent experiments (
14).
We believe that the key to achieving long-term culture of
T. pallidum in the current study is the combination of the use of the TpCM-2 modified medium and the maintenance of near-homeostatic conditions through regular subculture and partial medium replacement as needed. TpCM-2 differs from its precursor, TpCM, by substitution of Eagle’s MEM with CMRL 1066 medium as the basal medium and the omission of some of the TpCM additives (
Table 4). CMRL 1066 medium was selected because it is utilized as the basal medium in Barbour-Stoenner-Kelly (BSK) medium, which is widely used for culture of
Borrelia (relapsing fever) and
Borreliella (Lyme disease) isolates (
39,
40). Compared to MEM, CMRL contains additional nutrients, including nucleic acid bases, nucleosides, cocarboxylase, coenzyme A, flavin adenine dinucleotide, NAD, NADP, Na acetate, Na glucuronate, and the fatty acid mixture Tween 80; some of the shared components are also present at different concentrations. Further analysis will be needed to determine (i) if the CMRL-MEM substitution is required for long-term
T. pallidum culture and, if so, (ii) what CMRL component(s) is responsible for enhanced
T. pallidum survival and growth. The other factor that appears to be important in prolonged survival and growth is regular subculture at 6-to-7-day intervals. The relatively short time between subcultures limits both the depletion of nutrients and the development of toxic conditions, e.g., changes in pH and accumulation of toxic by-products. It should be noted, however, that the subculture interval can be increased when the number of
T. pallidum organisms in the culture is low and partial medium replacement is utilized. For example, in some of our low-inoculum experiments the level in the inocula became ≤10
6 per culture. To bolster these cultures, they were at times extended by up to 15 days with partial medium replacement at 4-to-7-day intervals (see, e.g.,
Fig. 1, low inoculum, days 59 to 74).
Another observation was that a maximum yield of
T. pallidum organisms per culture occurred in standard 9-cm
2 cultures with 2 ml of TpCM-2; the yield per culture was limited despite inoculum size when all other conditions are held constant. We found that a simple increase in the volume of medium resulted in a proportionately improved yield when a high inoculum was used (
Fig. 8). Therefore, it appears that
T. pallidum itself consumes nutrients and/or alters culture conditions in a way that limits its
in vitro growth, as is the case for most bacterial cultures. The yield of
T. pallidum organisms per culture could also be increased by using 75-cm
2 flasks and concomitant increases in inoculum size, medium volume, and Sf1Ep cell numbers (
Table S1).
Optimal growth of
T. pallidum in this system required the presence of Sf1Ep cells, with little multiplication occurring in parallel axenic cultures (
Fig. 3; see also
Fig. S3 in the supplemental material). It is of interest that the number of genome Equiv. per cell increased dramatically in axenic cultures, whereas this value remains relatively constant at ~3 genome Equiv. per cell in parallel cultures with Sf1Ep cells. Indeed, it had been observed previously that
T. pallidum in an axenic environment continues to synthesize DNA (and RNA) for up to 6 days, despite the fact that the number of cells did not increase in these experiments (
41,
42). Thus, the barrier to multiplication in the axenic cultures performed to date is not likely to be due to defects in genome replication. Direct interaction through adherence between Sf1Ep cells and
T. pallidum is apparently required for promotion of treponemal multiplication in that separation of
T. pallidum from the cell monolayer in Transwell chambers prevented growth and shortened survival (unpublished observations).
We speculate that
T. pallidum may directly acquire certain nutrients, such as lipids, through direct interaction with host cells. It had been shown previously that
B. burgdorferi can obtain lipids during direct interaction with mammalian cells (
43) and that
T. pallidum can acquire fluorescently labeled fatty acids directly from the surrounding medium (
44). Matthews et al. (
45) found that the lipid content of
T. pallidum purified from infected rabbit testes had high proportions of cholesterol, which was most likely acquired directly from the host rather than synthesized from fatty acids. These are important observations, since neither of these organisms contains the genes required for fatty acid synthesis (
46,
47).
T. pallidum, like other treponemes (
48–50), most likely acquires lipids bound to serum albumin, which acts as a detoxifying agent (
51); perhaps this mechanism is supplemented by direct acquisition from host cells. The bound particles or vesicles observed by cryo-electron microscopy (
Fig. 7; see also
Movies S1 and
S2 in the supplemental material) could conceivably represent a means of nutrient acquisition.
T. pallidum is a microaerophilic organism that requires low (1.5% to 5%) concentrations of oxygen for long-term survival and growth and yet is extremely sensitive to the toxic effects of atmospheric levels of oxygen (
16,
36,
42,
52–55). In 1974, Cox and Barber (
54) demonstrated that
T. pallidum consumed O
2, but even today we do not have a good explanation of how oxygen is utilized.
T. pallidum lacks genes encoding the tricarboxylic cycle, cytochromes, or other components of typical bacterial oxidative phosphorylation pathways (
47). One hypothesis is that O
2 is used as an electron acceptor to maintain appropriate NADH/NAD
+ levels through the action of NADH oxidase (
56).
T. pallidum lacks genes encoding superoxide dismutase or catalase but may utilize neelaredoxin to provide protective activity against superoxide and other reactive oxygen species (ROS) (
57). Reducing compounds such as DTT in TpCM and TpCM-2 scavenge ROS and play an important role in prolonging
T. pallidum survival
in vitro (
16). It is also possible that host cells
in vivo (and Sf1Ep cells in the
in vitro culture system) are active in scavenging of ROS and thus in protecting
T. pallidum from these toxic compounds.
Sf1Ep cells are better at supporting the survival and growth of
T. pallidum than are all other cell types that have been examined (
58). The particular capability of Sf1Ep cells to support
T. pallidum multiplication may be related to their relatively low growth rate and low metabolic activity. Sf1Ep cell cultures can survive quite well for 2 weeks, with little change in medium parameters such as pH. It is likely that other mammalian cell types deplete nutrients and change medium conditions more rapidly, limiting their efficacy in supporting
T. pallidum growth. We are hopeful that future studies will identify the nutrients or protective activities provided by Sf1Ep cells and thus permit long-term axenic survival and growth.
Retention of infectivity in the rabbit model and structural integrity are important indicators that in vitro-cultured T. pallidum organisms maintain wild-type properties. In future analyses, the genome sequence of long-term-cultured treponemes will be compared with those of rabbit-propagated organisms to determine whether any sequence differences are evident. However, selection of a small subset of variant cells capable of growing in vitro seems implausible, because exponential multiplication is evident within days after the inoculation of cultures with rabbit-derived treponemes.
The members of the
T. pallidum group of pathogens represent an extreme in terms of host dependence. These bacteria have adapted to survival only in mammalian tissue, relying on the host for provision of nucleic acid bases, fatty acids, most amino acids, and glucose as an energy source (
14,
47,
56) as well as for the maintenance of near-homeostatic conditions in terms of temperature, osmolarity, oxygen and CO
2 levels, and pH. The genus
Treponema is a genetically diverse group whose known members are primarily host-associated organisms, including commensal skin organisms, oral treponemes, intestinal spirochetes, oral- and hoof-associated organisms involved in polymicrobial infections, and termite gut symbionts (
59). However,
T. caldaria,
T. stenostreptum, and
T. zuelzerae were recently recognized as free-living
Treponema species (
60); many more environmental species are likely, as indicated by the identification of multiple
Treponema genomes in a microbiome study of water well sediment in Rifle, CO (
61). Progressive genome reduction is extreme in this genus, resulting in decreases from roughly 4 Mb for environmental and termite-associated treponemes to only 1.1 Mb for the
T. pallidum group, with intermediate genome sizes occurring in other mammal-associated
Treponema species (
18). Thus,
T. pallidum has lost genes for most biosynthetic pathways, stress response pathways, and complex energy production pathways while retaining the minimum complement of genes required for survival, proliferation, and transmission in the near-homeostatic environment of mammalian tissue, including those encoding transporters and efficient motility and chemotaxis systems.
In summary, the modification of the system used by Fieldsteel et al. (
15) described here will likely facilitate the characterization of the
T. pallidum subspecies,
T. carateum, and the rabbit and hare-infecting
Treponema species. Future studies will include application of this approach to culture of
T. pallidum subsp.
pertenue and subsp.
endemicum, which would in turn provide new avenues for the study of yaws and bejel. It is likely that the rabbit and hare pathogen (provisionally renamed “
T. paraluisleporidarum” from
T. paraluiscuniculi [
62]) can also be cultured by this method, which may provide new insights into the evolution of host specificity among
Treponema species. We anticipate that it will be possible to isolate
T. pallidum and other pathogenic
Treponema directly from tissues or body fluids using the
in vitro culture system. Potentially, this approach also may permit the culture of the pinta organism
T. carateum, which has been shown to be infectious in primates but has not been propagated in rabbits or other laboratory animals (
19,
27).
In vitro-cultured pathogenic treponemes could also be used in
in vitro host-pathogen interaction and immunologic studies, as well as in vaccine development. We have already begun to utilize the culture system for the following purposes: delineation of the nutrients required for
in vitro growth; comparison of the transcriptome and proteome of
in vitro-cultured
T. pallidum with those of rabbit-propagated organisms; cloning of pathogenic
Treponema by limiting dilution; genome sequencing of
in vitro-cultured
T. pallidum; and antimicrobial susceptibility testing. Random and targeted mutagenesis studies may also be possible using the
T. pallidum culture system. However, the current procedure is still complex, due to the requirement for tissue culture cells and a microaerobic environment. This aspect may hinder its widespread use by research laboratories, as well as its potential application to clinical uses. Therefore, the primary goal remains the development of an axenic system that supports the long-term culture of pathogenic
Treponema species, thereby further simplifying the study of these enigmatic organisms.