Borrelia burgdorferi infection-associated surface proteins ErpP, ErpA, and ErpC bind human plasminogen

Abstract

Host-derived plasmin plays a critical role in mammalian infection by Borrelia burgdorferi. The Lyme disease spirochete expresses several plasminogen-binding proteins. Bound plasminogen is converted to the serine protease plasmin and thereby may facilitate the bacterium's dissemination throughout the host by degrading extracellular matrix. In this work, we demonstrate plasminogen binding by three highly similar borrelial outer surface proteins, ErpP, ErpA, and ErpC, all of which are expressed during mammalian infection. Extensive characterization of ErpP demonstrated that this protein bound in a dose-dependent manner to lysine binding site I of plasminogen. Removal of three lysine residues from the carboxy terminus of ErpP significantly reduced binding of plasminogen, and the presence of a lysine analog, epsilon-aminocaproic acid, inhibited the ErpP-plasminogen interaction, thus strongly pointing to a primary role for lysine residues in plasminogen binding. Ionic interactions are not required in ErpP binding of plasminogen, as addition of excess NaCl or the polyanion heparin did not have any significant effect on binding. Plasminogen bound to ErpP could be converted to the active enzyme, plasmin. The three plasminogen-binding Erp proteins can also bind the host complement regulator factor H. Plasminogen and factor H bound simultaneously and did not compete for binding to ErpP, indicating separate binding sites for both host ligands and the ability of the borrelial surface proteins to bind both host proteins.

FIG. 1.

ErpA, ErpC, and ErpP bind plasminogen. Binding of plasminogen (25 μg/ml) to immobilized Erp proteins (10 μg/ml) was analyzed by ELISA, with bound plasminogen detected using specific antiserum. BSA was used as a negative control for nonspecific binding. Values represent plasminogen binding to each Erp protein minus background absorbance for BSA. Data represent the means and standard errors from two separate experiments with six replicates per Erp protein.

FIG. 2.

Alignment of ErpP, ErpA, and ErpC. Identical amino acid residues found in two or more of the proteins are boxed and shaded. The five amino acids deleted from the truncated ErpP used in this work are indicated by a thick line over the sequence. The black arrowhead indicates the first amino acid following the polyhistidine tag in the recombinant Erp protein.

FIG. 3.

ErpP binds plasminogen in a dose-dependent manner. Binding of ErpP (0 to 1 μM) to immobilized plasminogen (10 μg/ml) was analyzed by ELISA, with bound ErpP detected using polyclonal rabbit antiserum. BSA was used as a negative control for nonspecific binding. Values represent ErpP binding to plasminogen minus background absorbance for BSA. Data represent the means and standard errors from two separate experiments with six replicates per concentration of ErpP.

FIG. 4.

Role of lysines in ErpP plasminogen binding activity. Binding of plasminogen to immobilized ErpP (10 μg/ml) was analyzed by ELISA. (A) Plasminogen (25 μg/ml) was added to ErpP-coated wells in the presence or absence of 1 mM ɛ-aminocaproic acid (ACA). Bound plasminogen was detected using a specific antiserum. BSA was used as a negative control. Data represent the means and standard errors from two separate experiments with 12 replicates per condition. *, P < 0.001, Student's t test assuming unequal variances. (B) Binding of plasminogen to immobilized ErpP and the carboxy-terminal truncation mutant (rErpP2) was analyzed by ELISA. Plasminogen (25 μg/ml) was added to ErpP-coated wells. Bound plasminogen was detected using a specific antiserum. BSA was used as a negative control. Data represent the means and standard errors from two separate experiments with 12 replicates per condition. *, P < 0.001, Student's t test assuming unequal variances. (C) Binding of ErpP (0 to 250 nM) to immobilized LBS I of plasminogen analyzed by ELISA. Bound ErpP was detected using a specific antiserum. BSA was used as a negative control. Data represent the means and standard errors from two separate experiments with six replicates per concentration of ErpP.

FIG. 5.

Role of ionic interactions in ErpP binding of plasminogen. Binding of plasminogen to immobilized ErpP (10 μg/ml) was analyzed by ELISA. (A) Plasminogen (25 μg/ml) was incubated in the presence of increasing concentrations of NaCl. Bound plasminogen was detected using a specific antiserum. BSA was used as a negative control. Data represent the means and standard errors from two experiments with six replicates per concentration of NaCl. (B) Heparin (0 to 50 μM) was added to immobilized ErpP for 1 h prior to the addition of plasminogen. After a single wash with PBS-T, plasminogen (25 μg/ml) was incubated in the presence of heparin (0 to 50 μM). Bound plasminogen was detected using a specific antiserum. BSA was used as a negative control. Data represent the means and standard errors from two experiments with six replicates per concentration of heparin.

FIG. 6.

Factor H and plasminogen do not compete for binding to ErpP. Plasminogen (200 nM) binding to immobilized ErpP (10 μg/ml) was assayed in the presence of increasing concentrations of factor H. Both serum proteins were detected by individual, specific antisera. In normal human serum, the molar ratio of plasminogen to factor H is 1:1.4. Data represent the means and standard errors from two different experiments with six replicates per condition.

FIG. 7.

ErpP-bound plasminogen is converted into plasmin. ErpP-coated wells of microtiter plates were incubated with plasminogen, urokinase (uPA), and/or a plasmin-specific chromogenic substrate. Proteolytic activity was measured by absorbance at 450 nM. Data represent the means and standard errors from two different experiments with six replicates per condition.