Analysis of a unique interaction between the complement regulatory protein factor H and the periodontal pathogen Treponema denticola

Abstract

Treponema denticola, a spirochete associated with periodontitis, is abundant at the leading edge of subgingival plaque, where it interacts with gingival epithelia. T. denticola produces a number of virulence factors, including dentilisin, a protease which is cytopathic to host cells, and FhbB, a unique T. denticola lipoprotein that binds complement regulatory proteins. Earlier analyses suggested that FhbB specifically bound to factor H (FH)-like protein 1 (FHL-1). However, by using dentilisin-deficient mutants of T. denticola, we found that T. denticola preferentially binds FH and not FHL-1, and that FH is then cleaved by dentilisin to yield an FH subfragment of approximately 50 kDa. FH bound to dentilisin-deficient mutants but was not cleaved and retained its ability to serve as a cofactor for factor I in the cleavage of C3b. To assess the molecular basis of the interaction of FhbB with FH, mutational analyses were conducted. Replacement of specific residues in widely separated domains of FhbB and disruption of a central alpha helix with coiled-coil formation probability attenuated or eliminated FH binding. The data presented here are the first to demonstrate the retention at the cell surface of a proteolytic cleavage product of FH. The precise role of this FH fragment in the host-pathogen interaction remains to be determined.

FIG. 1.

Analysis of FH binding by purified r-FhbB and r-FhbB subfragments with a pull-down assay. Pull-down assays were performed as described in the text. r-FhbB proteins and r-CspA (a positive control for FH binding) were incubated with human serum or purified FH for these analyses. After recovery of the complexes, the samples were immunoblotted and bound FH was detected by screening immunoblots with anti-human FH antiserum. The specific recombinant protein used in each reaction and the FH source (human serum, purified FH, or no FH) are indicated above each lane. As a control for detection, an immunoblot strip of purified FH-FHL-1 was also included.

FIG. 2.

Demonstration of the cleavage of FH by the T. denticola protease dentilisin. In panel A, whole cells of T. denticola (Td) strain 35405 (wild type [wt]) and dentilisin mutants CCE and CKE were incubated with purified human FH, immunoblotted, and screened with anti-human FH antiserum. In panel B, the potential cleavage of FH by wild-type T. denticola 35405 and the dentilisin mutant (CCE) was assessed over time (in minutes as indicated). Aliquots were recovered at different time points, immunoblotted, and screened with anti-FH antiserum.

FIG. 3.

Analysis of the ability of cell-bound FH and FH subfragments to serve as a cofactor for the factor I-mediated cleavage of C3b. T. denticola strain 35405 and dentilisin mutants (CCE and CKE) were incubated with or without FH, as indicated above the lanes. After addition of C3b and factor I, the samples were immunoblotted and screened with anti-C3b antiserum. Additional controls in which various combinations of purified C3b, factor I, and/or FH were combined (without cells) were performed and analyzed as described above.

FIG. 4.

Analysis of truncated FhbB proteins for the ability to bind FH. All r-FhbB proteins were generated as N-terminally S-tagged fusions that lack the leader peptide sequence. The nomenclature assigned to each mutant is indicated and reflects the amino acid span present in the protein. Expression of r-FhbB proteins in E. coli was assessed with S protein-HRP conjugate as described in the text. FH binding by r-FhbB proteins was assessed by using the FH ALBI assay as described in the text.

FIG. 5.

Analysis of the binding of FH to r-FhbB proteins with site-directed mutations within the coiled-coil domain. Lysates of E. coli expressing r-FhbB with site-directed mutations (indicated above each lane) were immunoblotted and screened with S protein-HRP conjugate or used in an FH ALBI assay as described in the text. Computer-determined coiled-coil probabilities for the site-directed mutants are indicated in parentheses. These values are followed by a listing of the specific substitutions that were introduced into the proteins. The wild-type (wt) residue and position number are followed by the introduced substitution. Neg., negative control.

FIG. 6.

Identification of r-FhbB proteins generated by random mutagenesis with altered FH binding ability. Immunoblots of cell lysates from E. coli clones expressing randomly generated r-FhbB mutant proteins were screened with HRP-conjugated S protein or utilized in FH ALBI assays (as indicated). Analyses were performed as described in the text. Neg., negative control; wt, wild type.

FIG. 7.

Sequence analysis of FhbB variants generated by random mutagenesis. The alignment presents the amino acid sequences of the random mutant forms of FhbB. The numbering is based on the FhbB sequence of the T. denticola 35405 strain. Residues that differ from those of wild-type FhbB are indicated for each mutant, while residues identical to the wild-type sequence are indicated by periods. Since the r-FhbB random mutant proteins were generated without leader peptides, sequences corresponding to the leader peptides are not shown (these positions are indicated by ∼). To the right of each protein mutant sequence, the total number of amino acid changes is indicated in the Δ column. The results of the FH binding analyses with each recombinant protein are indicated as follows: +, binding detected; +/−, weak binding; −, no binding. The computer-identified coiled-coil region is underlined, and the specific positions of the a-g coiled-coil-associated heptad repeat are shown above the wild-type sequence.