Strain-specific variation of the decorin-binding adhesin DbpA influences the tissue tropism of the lyme disease spirochete

Abstract

Lyme disease spirochetes demonstrate strain- and species-specific differences in tissue tropism. For example, the three major Lyme disease spirochete species, Borrelia burgdorferi sensu stricto, B. garinii, and B. afzelii, are each most commonly associated with overlapping but distinct spectra of clinical manifestations. Borrelia burgdorferi sensu stricto, the most common Lyme spirochete in the U.S., is closely associated with arthritis. The attachment of microbial pathogens to cells or to the extracellular matrix of target tissues may promote colonization and disease, and the Lyme disease spirochete encodes several surface proteins, including the decorin- and dermatan sulfate-binding adhesin DbpA, which vary among strains and have been postulated to contribute to strain-specific differences in tissue tropism. DbpA variants differ in their ability to bind to its host ligands and to cultured mammalian cells. To directly test whether variation in dbpA influences tissue tropism, we analyzed murine infection by isogenic B. burgdorferi strains that encode different dbpA alleles. Compared to dbpA alleles of B. afzelii strain VS461 or B. burgdorferi strain N40-D10/E9, dbpA of B. garinii strain PBr conferred the greatest decorin- and dermatan sulfate-binding activity, promoted the greatest colonization at the inoculation site and heart, and caused the most severe carditis. The dbpA of strain N40-D10/E9 conferred the weakest decorin- and GAG-binding activity, but the most robust joint colonization and was the only dbpA allele capable of conferring significant joint disease. Thus, dbpA mediates colonization and disease by the Lyme disease spirochete in an allele-dependent manner and may contribute to the etiology of distinct clinical manifestations associated with different Lyme disease strains. This study provides important support for the long-postulated model that strain-specific variations of Borrelia surface proteins influence tissue tropism.

Figure 1. DbpA variants produced in B. burgdorferi promote distinct binding activities to decorin and dermatan sulfate.

Binding of radiolabeled B. burgdorferi ML23/pBBE22, ML23ΔdbpBA/pBBE22 (“Vector”), and the deletion strain bearing a plasmid encoding DbpA and DbpB (“pDbpBA”), DbpA_VS461ΔC11 (“ΔC11”), DbpA_VS461 (“VS461”), DbpA_PBr (“PBr”) or DbpA_N40-D10/E9 (“N40-D10/E9”), to decorin, dermatan sulfate (Derm SO₄), or the negative control chondroitin-6-sulfate (see Materials and Methods). The non-adherent B. burgdorferi strain B314 harboring the empty vector pJF21 (“B314/Vector”) was also included as negative control. The percentage of bound bacteria was determined by radioactive counts in bound bacteria normalized to the counts in the inoculum.Each bar represents the mean of four independent determinations ± SEM. Statistical significance was determined using the one-way ANOVA test. Significant (P<0.05) differences in binding relative to the dbpBA deletion strain (“*”), between two strains relative to each other (“#”), or relative to the dbpA_VS461ΔC11-complemented strain (“†”) are indicated.

Figure 2. DbpA variants promote distinct B. burgdorferi inoculation site colonization during early infection.

C3H/HeN mice infected with 10⁴ B. burgdorferi strain ML23/pBBE22 (“ML23/Vector”), dbpBA deletion strain ML23ΔdbpBA/pBBE22(“Vector”), or the deletion strain bearing a plasmid encoding the indicated variants were sacrificed at 3 days post-infection. Bacterial loads at the inoculation site were determined by qPCR. Data shown are the mean bacterial loads ± SEM of 10 mice per group. Statistical significance was determined using a one-way ANOVA test. Significant (P<0.05) differences in spirochete number relative to the dbpBA deletion strain (“*”), between two strains relative to each other (“#”), or relative to the dbpA_VS461ΔC11-complemented strain (“†”) are indicated. (n.d.): not determined.

Figure 3. DbpA variants promote distinct B. burgdorferi tissue colonization profiles at 28 days post-infection.

C3H/HeN mice infected with 10⁴ B. burgdorferi strain ML23/pBBE22 (“ML23/Vector”), dbpBA deletion strain ML23ΔdbpBA/pBBE22 (“Vector”), or the deletion strain bearing a plasmid encoding the indicated DbpA variants were sacrificed at 28 days post-infection. The bacterial loads at the inoculation site, ear, bladder, heart, knee, and tibiotarsus joint were determined by qPCR. Data shown are the mean bacterial loads ± SEM of 10 mice per group. Statistical significance was determined using a one-way ANOVA test. Significant (P<0.05) differences in spirochete number relative to the dbpBA deletion strain (“*”), between two strains relative to each other (“#”), or relative to the dbpA_VS461ΔC11-complemented strain (“†”), are indicated. (n.d.): not determined. These data are described comprehensively with other post-infection time points in Table S1.

Figure 4. Differences in tissue tropism promoted by DbpA variants are not a function of an adaptive immune response.

C3H/HeN-SCID mice infected with 10³ B. burgdorferi ML23/pBBE22 (“ML23/Vector”), ML23ΔdbpBA/pBBE22 (“Vector”), or ML23ΔdbpBA bearing a plasmid encoding the indicated DbpA variants/mutant were sacrificed at 28 days post-infection. Bacterial loads at the inoculation site, ear, bladder, heart, knee, and tibiotarsus joint were determined by qPCR. Data shown are the mean bacterial loads ± SEM of 10 mice per group. Statistical significance was determined using a one-way ANOVA test. Significant differences (P<0.05) in spirochete number relative to the ML23ΔdbpBA (“*”), between two strains relative to each other (“#”), or relative to the pdbpA_VS461ΔC11-complemented strain (“†”), are indicated.

Figure 5. DbpA variants produced by B. burgdorferi lead to differences in tissue inflammation in C3H/HeN-infected mice.

(A) C3H/HeN mice infected with 10⁴ B. burgdorferi strain ML23/pBBE22 (“ML23/Vector”), dbpBA deletion strain ML23ΔdbpBA/pBBE22 (“Vector”), or the deletion strain bearing a plasmid encoding the indicated DbpA variants were sacrificed at 28 days post-infection. Inflammation of the heart (upper panel) and joints (lower panel) were assessed using Hematoxylin and Eosin staining histopathology. The upper panel and lower panel indicate the higher-resolution (10×, bar  = 58 µm) and lower-resolution (4×, bar  = 141 µm) images. (B) To quantitate inflammation of heart and joint tissues, at least ten random sections from each infection group were scored on a scale of 0–3 for the severity of carditis and arthritis, as indicated in Material and Methods. Statistical significance was determined using a one-way ANOVA test. Data shown are the mean inflammation score ± SD of 5 mice per group. Statistical significance was determined using a one-way ANOVA test. Significant (P<0.05) differences in the inflammation score relative to the dbpBA deletion strain (“*”) or between two strains relative to each other (“#”), are indicated.