Microbes and Infection 4 (2002) 1133–1140

www.elsevier.com/locate/micinf

Review

The immune response to infection with Treponema pallidum,

the stealth pathogen
Juan C. Salazar a,b,*, Karsten R.O. Hazlett a, Justin D. Radolf a,c
a
Center for Microbial Pathogenesis, MC 3710, University of Connecticut Medical School, 263 Farmington Avenue, Farmington, CT 06030, USA
b
Department of Pediatrics, Connecticut Children’s Medical Center, Division of Pediatric Infectious Diseases,
282 Washington Street, Hartford, CT 06106, USA
c
Department of Medicine, University of Connecticut Health Center, 263 Farmington Avenue, Farmington, CT 06030, USA

Abstract

Cutaneous immunobiology and spirochetal molecular biology have allowed investigators to propose a conceptual framework for the
development of both the innate and adaptive immune response to Treponema pallidum infection. While some clinical manifestations can
be attributed to humoral responses, most can be attributed to a combination of local innate and adaptive cellular immunity. © 2002 Éditions
scientifiques et médicales Elsevier SAS. All rights reserved.
Keywords: Treponema pallidum; Syphilis; Spirochetes; Innate immunity; Adaptive immunity

1. Introduction syphilis (caused by T. pallidum subspecies pallidum), en-

demic syphilis (T. pallidum subspecies endemicum), yaws
Syphilis, a chronic inflammatory disorder caused by the (T. pallidum, subspecies pertenue) and pinta (T. carateum).
spirochetal pathogen Treponema pallidum, has been recog- Even though T. pallidum was identified microscopically
nized as a sexually transmitted disease for at least five early in the 20th century, the inability to culture the
centuries [1]. During that time, it has plagued both the bacterium in vitro has limited research on the disease and its
patient and the physician with its protean clinical manifes- etiologic agent [4]. Furthermore, the lack of a murine model
tations, serious late complications, and diagnostic dilem- has greatly retarded the ability of investigators to elucidate
mas. Over 12 million cases per year of infectious syphilis the complex host response to this pathogen. Despite these
are estimated to occur worldwide, the vast majority within major deficiencies in the research armamentarium, the
developing countries [2]. The World Health Organization application of modern molecular and immunological tech-
estimates that in some nations the syphilis rates among niques has yielded considerable information regarding the
commercial sex workers are between 23% and 32%. In biology of T. pallidum and the immunopathogenesis of
contrast, according to the Centers for Disease Control and venereal syphilis.
Prevention, the rates of primary and secondary syphilis in
the United States are at their lowest since reporting began in
1941. These unprecedented low rates, coupled with the 2. Clinical and histopathologic features of syphilis
concentration of the disease in a small number of geo-
graphic foci, have prompted the development of the Na- T. pallidum is an obligate pathogen of humans with no
tional Plan to Eliminate Syphilis from the United States [3]. known environmental or zoonotic reservoir. The vast major-
T. pallidum belongs to one of five genera within the order ity of cases are acquired through intimate contact, usually
Spirochaetales. The genera that are pathogenic for man sexual, with an actively infected partner. The infecting
include Leptospira (leptospirosis), Borrelia (Lyme disease organisms are thought to gain entry through intact mucosal
and relapsing fever) and Treponema. Four human diseases surfaces or microscopic skin abrasions. Before the onset of
are caused by members of the genus Treponema: venereal clinical manifestations, the bacterium begins to multiply
locally and disseminate from the local site through blood
* Corresponding author. Tel.: +1-860-545-9490; fax: +1-860-545-9371. vessels and lymphatics. Interjunctional penetration of vas-
E-mail address: jsalaza@ccmckids.org (J.C. Salazar). cular endothelium is believed to be the mechanism by which
© 2002 Éditions scientifiques et médicales Elsevier SAS. All rights reserved.
PII: S 1 2 8 6 - 4 5 7 9 ( 0 2 ) 0 1 6 3 8 - 6
1134 J.C. Salazar et al. / Microbes and Infection 4 (2002) 1133–1140

T. pallidum gains access to various tissues following he-

matogenous dissemination [5]. Similar in vitro studies also
suggest that activation of endothelial cells as they are
traversed by treponemes results in de novo expression
and/or upregulation of leukocyte adhesion molecules in-
volved in the recruitment of phagocytic cells ultimately
responsible for clearing the invader [6].
Genital ulcers (chancres), the most common manifesta-
tion of primary syphilis, become clinically apparent within
3–4 weeks following infection. Organisms that disseminate
from the primary site of infection may be trapped by
regional lymph nodes or spread hematogenously throughout
the body, setting the stage for secondary syphilis. Although
secondary syphilis is characterized most typically by a rash,
a remarkable variety of visceral complications also are
recognized. Malaise, low-grade fever and diffuse, painless
lymph node enlargement also are commonly observed
during secondary syphilis. The natural history of secondary
syphilis is to resolve spontaneously, giving rise to an
asymptomatic stage referred to as latent syphilis. Early
Fig. 1. T. pallidum ultrastructure. (A) Membrane architecture of T. palli-
latency (the first 4 years post-infection) is characterized by
dum subspecies pallidum, as demonstrated by freeze-fracture electron
recurrences of secondary syphilis. Thereafter, during late microscopy. The figure shows the paucity of OM transmembrane proteins.
latency, inflammatory foci may progress at sites of persis- (B) Proposed molecular architecture of T. pallidum. The OM is depicted as
tent infection, but the individual is no longer infectious. having rare transmembrane proteins, while major immunogens are lipo-
Classic studies demonstrate that 15–40% of untreated pa- proteins (LP1 and LP2) anchored to the periplasmic leaflet of the OM. The
bacterium is also shown with periplasmic endoflagellum (Ef) and peptido-
tients with latent syphilis develop recrudescent forms of the
glycan (pg)-cytoplasmic membrane (CM) complex. (Adapted with permis-
disease collectively referred to as tertiary syphilis [7,8]. The sion from: D.L. Cox, et al., Infect. Immun. 60 (3) (1992) 1076–1083 [10]).
tertiary stage may involve any organ system but primarily
affects the central nervous system (neurosyphilis), the heart microscopy. The spirochete consists of a central protoplas-
and vascular structures (aortitis/aneurysms) and skin and mic cylinder bounded by a cytoplasmic membrane, an
bones (gummatous syphilis). These complications can ulti- overlying thin layer of peptidoglycan, and an outer mem-
mately lead to death. brane (OM). Two or three flagella, the organelles of motility,
Cellular infiltrates composed of lymphocytes, macroph- originate at each end of the organism, entwine about the cell
ages, and plasma cells accompanied by vasculopathic body within the periplasmic space, and overlap in the
changes of varying severity are the sine qua non of middle. Unlike the OMs of Gram-negative bacteria, that of
syphilitic lesions in all stages of disease and the histopatho- T. pallidum lacks the potent proinflammatory glycolipid
logic correlates of organ system dysfunction [4,9]. A wide lipopolysaccharide (LPS) and is largely devoid of trans-
variety of histological patterns can be superimposed on membrane proteins (Fig. 1) [10,11]. Despite considerable
these common abnormalities. The frequent presence of efforts, these rare OM proteins and presumptive targets for
granulomata underscores the importance of cellular immune protective immunity have defied molecular characterization.
mechanisms in the local response to the bacterium. The A number of studies have shown that the spirochete’s major
granulomata of gummatous disease assume a necrotizing membrane immunogens are hydrophilic polypeptides teth-
character and are thought to represent an exaggerated ered by covalently bound N-terminal lipids to the periplas-
delayed type hypersensitivity response to treponemal anti- mic leaflet of the cytoplasmic membrane, a topology which
gens. renders them inaccessible to the high titers of antibodies
they induce during acquired infection [11]. This unusual
molecular architecture is thought to explain the poor anti-
3. T. pallidum ultrastructure and molecular biology genicity of motile (i.e., intact) treponemes during incubation
with syphilitic serum in vitro and the bacterium’s impres-
3.1. Pathogenicity by stealth sive capacity for immune evasion in vivo, hence its desig-
nation as a stealth pathogen [12].
T. pallidum is a helically shaped micro-aerophilic bacte-
rium varying from 6 to 20 µm in length and with a diameter, 3.2. The T. pallidum genomic sequence
0.10–0.18 µm, that places it below the resolution of con-
ventional light microscopy; the unstained bacterium can, The availability of the T. pallidum genomic sequence has
however, be visualized by dark-field or phase contrast provided investigators with a powerful tool for elucidating
 J.C. Salazar et al. / Microbes and Infection 4 (2002) 1133–1140 1135

the mechanisms by which the syphilis spirochete elicits and

evades host defenses [13]. The genome is a circular chro-
mosome of 1,138,006 base pairs and contains 1041 open
reading frames (ORFs). Biological functions have been
predicted for more than 50% of these ORFs. Of the
remaining ORFs, 17% match hypothetical proteins from
other species. Thus, T. pallidum contains a considerable
number of functionally uncharacterized ORFs that could
contribute to its distinctive and extremely complex parasitic
strategy. At the same time, consistent with biochemical
studies from the pre-genomic era, the bacterium is bereft of
numerous biosynthetic pathways found in cultivatable bac-
teria, emphasizing its absolute dependence upon the hostile
host milieu for sustenance. One particularly exciting dis-
covery is that the T. pallidum genome encodes 12 proteins
with varying degrees of sequence homology to the surface-
exposed major sheath protein of T. denticola, an oral
commensal. This finding has engendered speculation that
the syphilis spirochete possesses a system for antigenic
variation functionally analogous to those of relapsing fever
Fig. 2. Major membrane antigens of T. pallidum are lipoproteins. Immu-
spirochetes, and it has been evoked to explain the waxing
noblot analysis probed with human syphilitic serum of T. pallidum poly-
and waning character of syphilitic infection [14]. However, peptides after Triton X-114 phase partitioning of Percoll-purified orga-
it should be noted that evidence for such antigenic variation nisms. Lane 1, 3 × 108 whole, solubilized T. pallidum cells; lane 2, Triton
remains controversial; moreover, recent work by Hazlett et X-114-insoluble material; lane 3, detergent-phase proteins; and lane 4,
al. [15] has challenged this notion. A number of other aqueous-phase proteins. Major polypeptides detected in the detergent phase
(lane 3) are labeled a to g. Molecular size standards (in kilodaltons) listed
candidate OM proteins also have been “mined” from the
at the side. (Adapted with permission from: J.R. Radolf, et al., Infect.
genomic sequence, fueling optimism that bona fide rare OM Immun. 56 (1988) 490–498 [18]).
proteins will soon be identified and their vaccinogenic
potentials assessed.
patterns of antigenic reactivity in secondary and early latent
3.3. Genetic polymorphisms and strain subtyping syphilis tend to be highly similar from patient to patient. At
least 30 discrete antigens can be defined by immunoblotting
Although evidence from animal inoculation studies has of one-dimensional sodium dodecyl sulfate polyacrylamide
long pointed to the existence of antigenic differences among gel electrophoresis and over 60 are demonstrable by two-
T. pallidum clinical isolates (the so-called street strains), dimensional electrophoresis. Proteins with molecular
investigators have lacked the tools for differentiating masses of 47, 37, 35, 33, 30, 17 and 15 kDa are considered
strains. Using recently identified polymorphisms at two the major treponemal antigens and have been studied in
different genetic loci, Pillay et al. [16] have developed a recombinant form as potential diagnostic reagents (Fig. 2)
PCR-based assay for subtyping strains in clinical speci- [18].
mens. The resulting classification scheme should prove Although antibody binding to intact T. pallidum is diffi-
useful for epidemiological surveillance, vaccine studies, and cult to demonstrate microscopically, results from assays
assessment of the virulence potential of different isolates, utilizing motile organisms indicate that antibodies directed
issues which traditionally have been difficult for syphilis against treponemal surface antigens are produced during
researchers to address. infection: (a) passive administration of immune rabbit
serum to uninfected animals attenuates lesion development
following challenge [19]; (b) human syphilitic serum can
4. Systemic immune response to T. pallidum immobilize T. pallidum in the presence of complement
[20,21]; (c) infection-derived antibody dramatically en-
4.1. Humoral immunity hances phagocytosis of T. pallidum by macrophages in vitro
[22] (Fig. 3); and (d) attachment of T. pallidum to eukary-
Immunoblot analyses have shown that the number of otic cells can be blocked by patient sera [23]. The demon-
T. pallidum antigens recognized by syphilitic sera and the stration via freeze-fracture electron microscopy of aggre-
degree of reactivity is generally proportional to the duration gated particles in the OMs of treponemes incubated with
of clinical symptoms [17]. Except in early primary syphilis, immune sera further supports the contention that these
IgG titers far exceed those of IgM, although both classes of “functional” antibodies interact with surface-exposed anti-
immunoglobulins are produced throughout infection. The genic targets [24]. Taken as a whole, these studies suggest
1136 J.C. Salazar et al. / Microbes and Infection 4 (2002) 1133–1140

reactive lymphocytes are first seen in the spleen on days 3–6

following infections and maintain high antigen-specific
proliferative responsiveness from day 10 through 2 years of
infection. The time course for the development of this
response correlates well with the progression of mono-
nuclear cell infiltration of the primary site of infection (i.e.,
the testis) [25]. Specific reactive cells became more appar-
ent in the draining lymph nodes following bacterial clear-
ance during the resolution of the cellular infiltration. Cellu-
lar responses to the 37- and 30-kDa fractions were evident
by day 6 of infection, and responses to the 35-, 33-, and
14-kDa fractions were first detected on day 10; these
responses continued through the 7-month observation pe-
riod. In humans, appearance of T. pallidum-reactive cells in
peripheral blood occurs during late primary syphilis; the
highest responses occurred during the secondary stage, and
no response occurred in uninfected persons.

4.3. Protective immunity

It is well documented that patients treated for early

syphilis can be re-infected. However, there also is evidence
that immunity to exogenous infection (chancre immunity)
eventually develops in humans and rabbits, and such obser-
vations have been used to support the feasibility of a
syphilis vaccine. Controversial studies in humans [26]
demonstrated that individuals previously treated for syphilis
and challenged intradermally with T. pallidum (Nichols)
were less likely to develop a typical chancre at the site of
inoculation. A lack of clinical signs and/or symptoms was
most prominent in subjects with previously treated late
latent syphilis as compared to those with primary or
secondary disease. In rabbits, resistance to lesion develop-
ment following intradermal challenge becomes evident
Fig. 3. Immune rabbit serum enhances phagocytosis of T. pallidum (Tp) by about 3 weeks post-intratesticular inoculation, reaching a
macrophages (MΦ). Treponemes (10 organisms/macrophage) were incu- maximum about 12 weeks after the initial infection and
bated with rabbit peritoneal macrophages in the presence of 10% heat-
inactivated normal or immune rabbit sera. After incubation with T. palli-
remaining high for the remainder of the animal’s life.
dum, macrophages were washed, fixed and processed for Despite the lack of chancre development, these animals
immunofluorescence microscopy using human syphilitic sera and FITC- remain chronically infected with T. pallidum. In a landmark
conjugated goat anti-human IgG. Typical phagocytic vacuoles containing study, Miller [27] was able to induce complete protective
treponemal antigens are indicated by arrows. immunity in rabbits following repeated intravenous immu-
nization of rabbits with T. pallidum (Nichols) attenuated by
that biological assays provide a more sensitive readout of gamma-irradiation. To date this study remains the sole
surface antibody binding than microscopy-based method- example of complete immunity to T. pallidum challenge in
ologies. rabbits.
Nevertheless, there is still considerable controversy re-
4.2. T. pallidum-specific cellular responses garding the relative contributions of humoral and cellular
processes to the protective immune response. The ability of
Early studies of systemic cell-mediated responses to syphilitic sera to promote direct killing and opsonophago-
T. pallidum in humans often led investigators to conclude cytosis of virulent treponemes in vitro (per above) has been
that cellular immunity is suppressed during early syphilis. cited as evidence for a strong humoral component. This idea
This viewpoint, however, has been contradicted by more also is congruent with the generally accepted notion of
recent studies and is clearly at odds with the intense T. pallidum as an extracellular pathogen. On the other hand,
histological abnormalities that characterize syphilitic le- the development of acquired immunity during the latter
sions in both rabbits and humans. During the course of stages of syphilis when antibody titers are relatively low
primary testicular infection in the rabbit model, T. pallidum [26,28] argues in favor of a cellular component, as does the
 J.C. Salazar et al. / Microbes and Infection 4 (2002) 1133–1140 1137

failure of adoptively transferred immune reactive serum to that utilized by LPS [37,38]. This ostensible paradox was
completely protect naive rabbits from subsequent intrader- resolved by the discovery that native lipoproteins and
mal challenge with virulent T. pallidum [29]. Moreover, in synthetic analogs signal via toll-like receptor (TLR)2 as
his studies using gamma-irradiated organisms, Miller noted opposed to LPS-mediated signaling which proceeds via
that animals remained fully protected 1 year following the TLR4 [39–41].
completion of immunization in the face of generally low It must be acknowledged that the ultrastructural studies
anti-treponemal antibody titers and, in most cases, nonreac- described earlier raise an obvious paradox: if treponemal
tive T. pallidum-immobilizing antibody tests [27]. The dif- lipoproteins are entirely subsurface, how can they interact
ficulties in resolving these contradictory findings and view- with pattern recognition receptors (PRRs) on the surface of
points are an unfortunate byproduct of the lack of an inbred monocytes/macrophages? Indeed, virulent treponemes are
experimental animal model. extremely poor activators of cultured monocytic cells com-
pared to either T. pallidum lysates or live Borrelia burgdor-
feri, which possesses surface-exposed lipoproteins [42].
5. Tissue-based inflammatory response to T. pallidum: Two potential explanations, which are not mutually exclu-
innate and adaptive immunity sive, can be envisioned. The first is that the innate immune
responses are activated when lipoproteins released from
In recent years, there has been a growing appreciation dead organisms or circulating in the form of immune
that tissue-based inflammation during syphilis is fundamen- complexes bind to PRRs. The second, which we favor, is
tal to both disease progression (i.e., lesion formation) and that lipoproteins gain access to PRRs following degradation
bacterial clearance (i.e., lesion resolution) and that the of treponemes within phagolysosomal vacuoles. Three lines
presence of spirochetes within tissues constitutes the driving of evidence support this idea: (a) lipoproteins retain their
force for the activation of resident immune cells and the biological activities even after extensive proteolytic degra-
recruitment of immune effector cells from peripheral blood. dation [34]; (b) TLR2 molecule lines the phagocytic vacu-
Two developments have markedly enhanced our ability to oles of murine macrophages and is capable of “sampling”
examine these interrelated processes. The first is the discov- vacuolar contents [43]; and (c) treponemes become highly
ery that treponemal lipoproteins are the dominant proin- stimulatory for human monocytes when they are “driven”
flammatory agonists of syphilitic infection. The second is into phagolysosomes following incubation with the opsonic
the recognition that host defenses against bacterial patho- antibodies present in human syphilitic sera (unpublished
gens involve a complex interplay of innate and adaptive observations).
immune responses [30].
5.2. Treponemal lipoproteins elicit a complex cellular
5.1. Treponemal lipoproteins as proinflammatory agonists response in vivo

The notion that T. pallidum lipoproteins are proinflam- Can treponemal lipoproteins elicit an immune response
matory agonists in syphilis arose from the convergence of in vivo and, if so, what is the nature of the response? To
two separate lines of investigation. The first involved efforts address this critical question, we first used the rabbit model
by us and others to clarify the syphilis spirochete’s distinc- to demonstrate that intradermal injection of lipopeptides
tive molecular architecture; as noted already, these studies elicits cellular infiltrates resembling those observed in
led to the discovery that the major membrane immunogens acquired syphilis in humans [44]. A detailed analysis of the
of T. pallidum are lipoproteins [11]. The second was the infiltrating cells was precluded, however, by the paucity of
pioneering studies by Bessler and Jung [31–33] demonstrat- reagents directed against rabbit immune cells and immuno-
ing that murein lipoprotein of Escherichia coli and synthetic modulators. We therefore turned to an in vivo human skin
lipopeptides based upon the molecule’s N terminus have the model in order to characterize further the in vivo biological
capacity to activate monocytes/macrophages and B cells. properties of T. pallidum lipoproteins [45]. In this model,
Knowing that T. pallidum lacks LPS, the Bessler and Jung mild vacuum suction is used to elicit blister formation at the
studies prompted us to investigate whether treponemal dermo-epidermal junction following intradermal injection
lipoproteins possess proinflammatory properties compa- with treponemal lipopeptides (Fig. 4); fluid within the
rable to those of murein lipoprotein. Indeed, using both blisters contains extravasated immune cells which can then
purified treponemal lipoproteins and synthetic lipopeptides, be analyzed by flow cytometry and compared to peripheral
this conjecture was confirmed in a series of in vitro studies blood. Prior to this study there already existed extensive
which also demonstrated that these activities were abso- evidence that the cellular composition of blister fluid
lutely dependent upon the presence of covalently bound accurately reflects the cellular infiltrates within the under-
N-terminal lipids [34–36]. Subsequent analyses revealed lying dermis.
that as with LPS, cellular activation by lipopro- Preliminary studies revealed that lipopeptides derived
teins/lipopeptides was facilitated by CD14 and NF-jB- from the 17- and 47-kDa T. pallidum lipoproteins elicited
dependent, but involved a signaling pathway distinct from dose-dependent gross inflammatory responses which
1138 J.C. Salazar et al. / Microbes and Infection 4 (2002) 1133–1140

Fig. 4. Epidermal blisters. A suction blister apparatus and epidermal

blisters (inset) elicited after 2 h of vacuum at 200 mm Hg. Blister
formation occurs at the epidermal–dermal interface.

peaked approximately 48 h following injection (Fig. 5). tion of the CD4+ cells in secondary lesions are actually
Consistent with previously published in vitro studies, the macrophages and that macrophages, not T cells, are the
corresponding nonlipidated peptides failed to elicit discern- predominant cell type [9]. Analysis of T-cell subsets indi-
ible gross inflammatory responses. Of particular impor- cated that the relative proportion of CD4 and CD8 cells
tance, blister fluids raised over the lipopeptide injection sites shifts from a CD4 predominance in chancres to a CD8
typically contained hundreds of thousands of leukocytes, majority in secondary lesions [9,46]. The presence of
whereas blisters raised over sites injected with peptides cytotoxic T cells in substantial numbers is of interest given
contained only a modicum of cells. Compared with periph- that one does not usually associate this cell type with the
eral blood, lipopeptide blister fluids were highly enriched immune response to an extracellular pathogen. Cells infil-
for monocytes/macrophages, cutaneous lymphocyte antigen trating primary and secondary syphilis lesions contain
(CLA)-positive T cells, and dendritic cells. The enrichment mRNA for the Th1 cytokines, interleukin 2 (IL-2), gamma
for CLA-positive T cells was of particular interest given that interferon (IFN γ), and IL-12 [47]. Increased expression of
cells bearing this skin addressin typically give rise to a Th1 IL-2 and IFN γ mRNAs by infiltrating cells also has been
response. Subsequent unpublished studies have shown that demonstrated in the rabbit model [48]. It is evident that
T cells recruited by treponemal lipopeptides express other these studies have provided only a bare outline of the local
surface markers indicative of a Th1 phenotype. Staining for adaptive response and that much remains to be learned
maturation/differentiation markers (CD83, CD1a), costimu-
latory molecules (CD80/CD86), and chemokine receptors
(CXCR4 and CCR5) revealed that blister fluid dendritic
cells, particularly the monocytoid subpopulation, were
highly activated and far more developmentally advanced
than their peripheral blood counterparts. Taken as a whole,
these studies demonstrated that treponemal lipopeptides
have the capacity to induce an inflammatory milieu remi-
niscent of that found in early syphilis lesions and that, in a
tissue environment, these agonists recruit a cellular infiltrate
with the potential to bridge innate and adaptive immunity.

5.3. Adaptive immunity

Investigators have used immunocytochemical techniques Fig. 5. Immunophenotypic characterization of leukocyte populations in
and, more recently, RT-PCR to probe biopsy specimens blister fluids and whole blood. The lipopeptide-induced proinflammatory
from primary and secondary syphilitic lesions for the response (% live cells) in blister fluid (BF-17L and BF-47L) is shown in
components of adaptive immunity. These studies have comparison to the response in peripheral blood (PB). Asterisks indicate
values that are significantly different than the corresponding value in whole
revealed, not surprisingly, that the cellular infiltrates are blood (P < 0.01). Cells in BF and PB were stained with FITC CD14, PE
composed predominantly of lymphocytes and macrophages. CD20, PerCP CD45, and APC CD38 and analyzed by flow cytometric
Interestingly, our studies indicated that a substantial propor- analysis.
 J.C. Salazar et al. / Microbes and Infection 4 (2002) 1133–1140 1139

regarding the immunophenotypes and antigenic specificities [3] M.E. St Louis, J.N. Wasserheit, Elimination of syphilis in the United
of the infiltrating T cells. States, Science 281 (5375) (1998) 353–354.
[4] S.A. Lukehart, K.K. Holmes, Syphilis, in: E. Braunwald, K.J. Issel-
bacher, R.G. Petersdorf, J.D. Wilson, J.B. Martin (Eds.), Harrison’s
Principles of Internal Medicine, McGraw Hill Book Company, New
6. A unifying hypothesis York, 1998, pp. 1023–1033.
[5] D.D. Thomas, M. Navab, D.A. Haake, A.M. Fogelman, J.N. Miller,
Based upon the findings discussed above, one can pro- M.A. Lovett, Treponema pallidum invades intercellular junctions of
endothelial cell monolayers, 85, Proc. Natl. Acad. Sci., USA, 1988,
pose a scenario for the ontogeny of cellular immunity to pp. 3608–3612.
T. pallidum that integrates innate and adaptive responses. [6] B.S. Riley, N. Oppenheimer-Marks, E.J. Hansen, J.D. Radolf,
According to this scheme, pre-programmed innate mecha- M.V. Norgard, Virulent Treponema pallidum activates human vas-
nisms are activated by treponemal lipoproteins in all in- cular endothelial cells, J. Infect. Dis. 165 (1992) 484–493.
fected tissues and in all stages of disease; this “lipoprotein [7] E.G. Clark, N. Danbolt, The, Oslo study of the natural course of
untreated syphilis. An epidemiologic investigation based on a
effect” can be invoked to explain the common histopatho-
re-study of the Boeck–Bruusgaard material, Med. Clin. North Am.
logical features of syphilis lesions and the predominance of 48 (1964) 613–623.
macrophages [4]. Superimposed on this proinflammatory [8] S.R. Thomas, S.C. Quinn, The Tuskegee syphilis study, 1932 to
“backdrop” is an increasingly robust adaptive (i.e., 1972: implications for HIV education and AIDS risk education
T. pallidum-specific) response. Early in the development of programs in the black community, Am. J. Public Health 81 (1991)
1498–1505.
a chancre, a primary immune response ensues because the
[9] R.L. McBroom, A.R. Styles, M.J. Chiu, C. Clegg, C.J. Cockerell,
infiltrating memory effector T cells are not sensitized to J.D. Radolf, Secondary syphilis in persons infected with and not
treponemal antigens. As a result, antigen-presenting cells infected with HIV-1: a comparative immunohistological study, Am.
(APCs) attracted to the site of inflammation must traffic to J. Dermatopathol. 21 (1999) 432–441.
draining lymph nodes in order to present treponemal anti- [10] D.L. Cox, P. Chang, A. McDowall, J.D. Radolf, The outer mem-
brane, not a coat of host proteins, limits the antigenicity of virulent
gens to naive T cells. Trafficking of sensitized T cells back
Treponema pallidum, Infect. Immun. 60 (3) (1992) 1076–1083.
to the genital ulcer would be expected to accelerate bacterial [11] J.D. Radolf, Treponema pallidum and the quest for outer membrane
clearance and subsequent healing of the primary lesion. In proteins, Mol. Microbiol. 16 (1995) 1067–1073.
secondary syphilis lesions, the influx of pre-existing [12] J.D. Radolf, Role of outer membrane architecture in immune
T. pallidum-sensitized T cells would foster a response that evasion by Treponema pallidum and Borrelia burgdorferi, Trends
more efficiently contains local infection, most likely in Microbiol. 2 (1994) 307–311.
[13] C.M. Fraser, S.J. Norris, G.M. Weinstock, O. White, G.G. Sutton,
conjunction with opsonizing antibodies. With progressive R. Dodson, et al., Complete genome sequence of Treponema
T-cell sensitization (again, most likely in conjunction with pallidum, the syphilis spirochete, Science 281 (5375) (1998)
the production of opsonizing antibodies), the host immune 375–388.
response gains the upper hand, secondary relapses become [14] A. Centurion-Lara, C. Castro, L. Barrett, C. Cameron, M. Mostowfi,
less frequent, and the patient progresses from early to late W.C. Van Voorhis, et al., Treponema pallidum major sheath protein
homolog Tpr K is a target of opsonic antibody and the protective
latency. An age-related breakdown of a critical surveillance
immune response, J. Exp. Med. 189 (1999) 647–656.
function in a highly sensitized, persistently infected host [15] K.R.O. Hazlett, T.J. Sellati, T.T. Nguyen, D.L. Cox, M.L. Clawson,
would then give rise to recrudescent disease. M.J. Caimano, et al., The TprK protein of Treponema pallidum is
periplasmic and is not a target of opsonic antibody or protective
immunity, J. Exp. Med. 193 (2001) 1015–1026.
[16] A. Pillay, H. Liu, C.Y. Chen, B. Holloway, A.W. Sturm,
Acknowledgements
B. Steiner, et al., Molecular subtyping of Treponema pallidum
subspecies pallidum, Sex Transm. Dis. 25 (1998) 408–414.
This work was supported in part by Public Health Service [17] S.A. Lukehart, Immunology and pathogenesis of syphilis, in:
Grants AI-26756 (to JDR), AI-38894 (to JDR) and AI- T.C. Quinn (Ed.), Advances in Host Defense Mechanisms. Sexually
10573 (to KROH), General Clinical Research Center Transmitted Diseases, Raven Press, New York, 1992, pp. 141–163.
(GCRC) Grant M01RR06192 awarded to the University of [18] J.D. Radolf, N.R. Chamberlain, A. Clausell, M.V. Norgard, Identi-
fication and localization of integral membrane proteins of Tre-
Connecticut Health Center by the National Institutes of ponema pallidum subsp. pallidum by phase partitioning with the
Health, and The Donaghue Medical Research Foundation nonionic detergent Triton X-114, Infect. Immun. 56 (1988) 490–498.
Grant DF 00-014 (to JCS). We are indebted to Ken Bourell [19] N.H. Bishop, J.N. Miller, Humoral immunity in experimental
for his support with graphic design. syphilis. I. The demonstration of resistance conferred by passive
immunization, J. Immunol. 117 (1976) 191–196.
[20] R.A. Nelson Jr, M.M. Mayer, Immobilization of Treponema palli-
dum in vitro by antibody produced in syphilitic infection, J. Exp.
References Med. 89 (1949) 369–383.
[21] N.H. Bishop, J.N. Miller, Humoral immunity in experimental
[1] C.C. Dennie, A History of Syphilis, in: C. Charles (Ed.), Thomas, syphilis. II. The relationship of neutralizing factors in immune sera
Springfield, 1962. to acquired resistance, J. Immunol. 117 (1976) 197–207.
[2] A.E. Singh, B. Romanowski, Syphilis: review with emphasis on [22] S.A. Lukehart, J.N. Miller, Demonstration of the in vitro phagocy-
clinical, epidemiologic, and some biologic features, Clin. Microbiol. tosis of Treponema pallidum by rabbit peritoneal macrophages,
Rev. 12 (2) (1999) 187–209. J. Immunol. 121 (1978) 2014–2024.
1140 J.C. Salazar et al. / Microbes and Infection 4 (2002) 1133–1140

[23] N.S. Hayes, K.E. Muse, A.M. Collier, J.B. Baseman, Parasitism by [37] M.V. Norgard, L.L. Arndt, D.R. Akins, L.L. Curetty, D.A. Harrich,
virulent Treponema pallidum of host cell surfaces, Infect. Immun. 17 J.D. Radolf, Activation of human monocytic cells by Treponema
(1977) 174–186. pallidum and Borrelia burgdorferi lipoproteins and synthetic li-
[24] D.R. Blanco, J.D. Radolf, M.A. Lovett, J.N. Miller, The antigenic popeptides proceeds via a pathway distinct from that of lipopolysac-
interrelationship between the endoflagella of Treponema phagedenis charide but involves the transcriptional activator NF-kappa B,
biotype Reiter and Treponema pallidum, Nichols strain. I. Trepone- Infect. Immun. 64 (9) (1996) 3845–3852.
micidal activity of cross-reactive endoflagellar antibodies against T. [38] T.J. Sellati, D.A. Bouis, R.L. Kitchens, R.P. Darveau, J. Pugin,
pallidum, J. Immunol. 137 (1986) 2973–2979. R.J. Ulevitch, et al., Treponema pallidum and Borrelia burgdorferi
[25] S.A. Lukehart, S.A. Baker-Zander, R.M. Cheri Lloyd, S. Sell, lipoproteins and synthetic lipopeptides activate monocytic cells via
Characterization of lymphocyte responsiveness in early experimen- a CD14-dependent pathway distinct from that utilized by li-
tal syphilis. II. Nature of cellular infiltration and Treponema palli- popolysaccharide, J. Immunol. 160 (1998) 5455–5464.
dum distribution in testicular lesions, J. Immunol. 124 (1980) [39] E. Lien, T.J. Sellati, A. Yoshimura, T.H. Flo, G. Rawadi, R.W. Fin-
461–467. berg, et al., Toll-like receptor 2 functions as a pattern recognition
[26] H.J. Magnuson, E.W. Thomas, S. Olansky, B.I. Kaplan, L. de Mello, receptor for diverse bacterial products, J. Biol. Chem. 274 (47)
J.C. Cutler, Inoculation syphilis in human volunteers, Medicine 35 (1999) 33419–33425.
(1956) 33–82.
[40] H.D. Brightbill, D.H. Libraty, S.R. Krutzik, R.B. Yang, J.T. Belisle,
[27] J.N. Miller, Immunity in experimental syphilis VI. Successful
J.R. Bleharski, et al., Host defense mechanisms triggered by
vaccination of rabbits with Treponema pallidum, Nichols strain,
microbial lipoproteins through toll-like receptors, Science 285
attenuated by gamma-irradiation, J. Immunol. 110 (1973)
(1999) 732–736.
1206–1215.
[28] P.A. Hanff, T.E. Fehniger, J.N. Miller, M.A. Lovett, Humoral [41] A. Poltorak, X. He, I. Smirnova, M.Y. Liu, C. Van Huffel,
immune response in human syphilis to polypeptides of Treponema X. Du, et al., Defective LPS signaling in C3H/HeJ and
pallidum, J. Immunol. 129 (1982) 1287–1291. C57Bl/10ScCr mice: mutations in Tlr4 gene, Science 282 (1998)
[29] N.H. Bishop, J.N. Miller, Humoral immune mechanisms in experi- 2085–2088.
mental syphilis, in: R.F. Schell, D.M. Musher (Eds.), Pathogenesis [42] T.J. Sellati, D.A. Bouis, M.J. Caimano, A.J. Feulner, C. Ayers,
and Immunology of Treponemal Infection, Marcel Dekker, New E. Lien, et al., Activation of human monocytic cells by Borrelia
York, 1983, pp. 241–270. burgdorferi and Treponema pallidum is facilitated by CD14 and
[30] D.T. Fearon, R.M. Locksley, The instructive role of innate immunity correlates with surface exposure of spirochetal lipoproteins, J. Im-
in the acquired immune response, Science 272 (1996) 50–53. munol. 163 (1999) 2049–2056.
[31] W.G. Bessler, M. Cox, A. Lex, B. Suhr, K.H. Wiesmuller, G. Jung, [43] D.M. Underhill, A. Ozinsky, A.M. Hajjar, A. Stevens, C.B. Wilson,
Synthetic lipopeptide analogs of bacterial lipoprotein are potent M. Bassetti, et al., The Toll-like receptor 2 is recruited to macroph-
polyclonal activators for murine B lymphocytes, J. Immunol. 135 age phagosomes and discriminates between pathogens, Nature 401
(1985) 1900–1905. (6755) (1999) 811–815.
[32] S. Hauschildt, P. Hoffmann, H.U. Beuscher, G. Dufhues, P. Heinrich, [44] M.V. Norgard, B.S. Riley, J.A. Richardson, J.D. Radolf, Dermal
K.H. Wiesmuller, et al., Activation of bone marrow-derived mouse inflammation elicited by synthetic analogs of Treponema pallidum
macrophages by bacterial lipopeptide: cytokine production, phago- and Borrelia burgdorferi lipoproteins, Infect. Immun. 63 (1995)
cytosis, and Ia expression, Eur. J. Immunol. 20 (1990) 63–68. 1507–1515.
[33] P. Hoffmann, S. Heinle, U.F. Schade, H. Loppnow, A.J. Ulmer,
[45] T.J. Sellati, S.L. Waldrop, J.C. Salazar, P.R. Bergstresser, L.J. Picker,
H.D. Flad, et al., Stimulation of human and murine adherent cells by
J.D. Radolf, The cutaneous response in humans to Treponema
bacterial lipoprotein and synthetic lipopeptide analogs, Immunobi-
pallidum lipoproteins involves cellular elements of both innate and
ology 177 (1988) 158–170.
adaptive immunity, J. Immunol. 166 (2001) 4131–4140.
[34] J.D. Radolf, M.V. Norgard, M.E. Brandt, R.D. Isaacs, P.A. Thomp-
son, B. Beutler, Lipoproteins of Borrelia burgdorferi and Treponema [46] A. Tosca, J. Lehou, M. Hatjivasiliou, A. Varelzidis, J.D. Stratigos,
pallidum activate cachectin/tumor necrosis factor synthesis: analysis Infiltrate of syphilitic lesions before and after treatment, Genitourin.
using a CAT reporter construct, J. Immunol. 147 (1991) 1968–1974. Med. 64 (1988) 289–293.
[35] D.R. Akins, B.K. Purcell, M. Mitra, M.V. Norgard, J.D. Radolf, [47] W.C. Van Voorhis, L.K. Barrett, D.M. Koelle, J.M. Nasio,
Lipid modification of the 17-kilodalton membrane immunogen of F.A. Plummer, S.A. Lukehart, Primary and secondary syphilis
Treponema pallidum determines macrophage activation as well as lesions contain mRNA for Th1 cytokines, J. Infect. Dis. 173 (1996)
amphiphilicity, Infect. Immun. 61 (1993) 1202–1210. 491–495.
[36] J.D. Radolf, L.L. Arndt, D.R. Akins, L.L. Curetty, M.E. Levi, [48] T.W. Arroll, A. Centurion-Lara, S.A. Lukehart, W.C. Van Voorhis,
Y. Shen, et al., Treponema pallidum and Borrelia burgdorferi T-Cell responses to Treponema pallidum subsp. pallidum antigens
lipoproteins and synthetic lipopeptides activate during the course of experimental syphilis infection, Infect. Immun.
monocytes/macrophages, J. Immunol. 154 (1995) 2866–2877. 67 (1999) 4757–4763.