The purpose of this investigation was to use molecular and immunological techniques to determine whether oral Treponema infected the human brain. Pieces of frontal lobe cortex from 34 subjects were analyzed with species-specific PCR and monoclonal antibodies. PCR detected Treponema in 14/16 Alzheimer's disease (AD) and 4/18 non-AD donors (P < 0.001), and AD specimens had more Treponema species than controls (P < 0.001). PCR also detected Treponema in trigeminal ganglia from three AD and two control donors. Cortex from 15/16 AD subjects and 6/18 controls contained Treponema pectinovorum and/or Treponema socranskii species-specific antigens (P < 0.01). T. pectinovorum and/or T. socranskii antigens were also found in trigeminal ganglia and pons from four embalmed cadavers, and 2/4 cadavers also had Treponema in the hippocampus. These findings suggest that oral Treponema may infect the brain via branches of the trigeminal nerve.

Species-specific nested PCR was used to detect Treponema amylovorum, Treponema denticola, Treponema maltophilum, Treponema medium, Treponema pectinovorum, Treponema socranskii, andTreponema vincentii in dental plaque. Subjects with periodontitis harbored all species, but T. pectinovorum andT. vincentii were not found in plaque from disease-free subjects.

Recent findings challenge the assumption that pathogen-related oral spirochetes (PROS) are related to Treponema pallidum. Treponema vincentii, grown in OMIZ-Pat media, cross-reacted with monoclonal antibody H9-2 against T. pallidum, and cultivable PROS had 16S rRNA gene sequences similar to those of T. vincentii(C.-B. Choi, C. Wyss, and U. B. Göbel. J. Clin. Microbiol. 34:1922–1925, 1996). Aims of the present study were to determine whether antigen phenotypes of oral treponemas were influenced by growth conditions and to evaluate the genetic relatedness of cultivable PROS to T. pallidum and T. vincentii. Results show that three T. pallidummonoclonal antibodies (H9-1, H9-2, and F5) cross-reacted with whole cells from four Treponema species grown in modified OMIZ-Pat medium, but not with treponemas grown in NOS medium. Only H9-2 reacted in immunoblots with reduced proteins from cultivable PROS andT. vincentii. Three of five PROS isolates were amplified byT. vincentii-specific PCR, and one was amplified byTreponema medium-specific PCR. None were amplified byT. pallidum-specific PCR. Three of five PROS isolates had 16S ribosomal DNA restriction fragment length polymorphism patterns identical to that of T. vincentii, and the patterns of two isolates resembled that of T. medium. Arbitrarily primed-PCR profiles from whole genomic DNA were distinct among five PROS isolates and two T. vincentii strains. Thus, PROS isolates represent a heterogeneous group of treponemas that share some 16S rRNA gene sequences with T. vincentii and T. medium, but not with T. pallidum. It is proposed that the PROS nomenclature be dropped.