T Cells Exacerbate Lyme Borreliosis in TLR2-Deficient Mice

Abstract

Infection of humans with the spirochete, Borrelia burgdorferi, causes Lyme borreliosis and can lead to clinical manifestations such as arthritis, carditis, and neurological conditions. Experimental infection of mice recapitulates many of these symptoms and serves as a model system for the investigation of disease pathogenesis and immunity. Innate immunity is known to drive the development of Lyme arthritis and carditis, but the mechanisms driving this response remain unclear. Innate immune cells recognize B. burgdorferi surface lipoproteins primarily via toll-like receptor (TLR)2; however, previous work has demonstrated TLR2^-/- mice had exacerbated disease and increased bacterial burden. We demonstrate increased CD4 and CD8 T cell infiltrates in B. burgdorferi-infected joints and hearts of C3H TLR2^-/- mice. In vivo depletion of either CD4 or CD8 T cells reduced Borrelia-induced joint swelling and lowered tissue spirochete burden, whereas depletion of CD8 T cells alone reduced disease severity scores. Exacerbation of Lyme arthritis correlated with increased production of CXCL9 by synoviocytes, and this was reduced with CD8 T cell depletion. These results demonstrate T cells can exacerbate Lyme disease pathogenesis and prolong disease resolution possibly through dysregulation of inflammatory responses and inhibition of bacterial clearance.

Keywords: Borrelia burgdorferi; Lyme disease; T cells; arthritis; mouse; toll-like receptor 2.

Figure 1

Arthritis development and T cell infiltrates in TLR2^−/− mice. (A) Ankle diameters of WT and TLR2^−/− mice were measured weekly following footpad inoculation with 5 × 10⁴ B. burgdorferi. At each indicated time point, three mice per strain were sacrificed and numbers of CD3e⁺ T cells were determined from (B) ankles or (C) hearts by flow cytometry. Data are representative of two separate experiments. Symbols represent means ± SD. *p > 0.01 compared to WT control from same time point.

Figure 2

T cell subsets in infected ankle and heart tissue. WT and TLR2^−/− mice were infected with B. burgdorferi and three mice from each strain sacrificed at the indicated time points. T cell subsets in ankle (A,C) and heart (B,D) tissue were characterized for CD4 (A,B), CD8 (C,D) by flow cytometry. Data are representative of two separate experiments. Bars represent means + SD. *p > 0.05 compared to WT control from same time point.

Figure 3

NK and NKT cells in infected ankle and heart tissue. WT and TLR2^−/− mice were infected with B. burgdorferi and three mice from each strain sacrificed at the indicated time points. NK (CD45.2⁺NKp46⁺CD122⁺CD3e⁻) and NKT (CD45.2⁺NKp46⁺CD122⁺CD3e⁺) cells were enumerated in ankle (A,B) and heart (C,D) tissue at day 21 postinfection. Data are representative of two separate experiments. Bars represent means + SD. *p > 0.05 compared to WT control from same time point.

Figure 4

Lyme arthritis in mice depleted of CD4 or CD8 T cells. Mice were treated with saline (control), CD4-depleting antibody, or CD8-depleting antibody and infected with B. burgdorferi. Flow cytometry plots demonstrate depletion of CD4⁺ and CD8⁺ T cells 6 days posttreatment (A). Arthritis development was monitored over time for WT (B) and TLR2^−/− (C) mice. Representative pictures of TLR2^−/− mice depleted of CD8 T cells or saline treated were taken at day 14 postinfection (D). Data are representative of two separate experiments. n = 3, symbols represent means ± SD. *p > 0.01 compared to WT control from same time point.

Figure 5

Arthritis and carditis severity scores. Mice were infected and treated as in Figure 4. On day 21 postinfection, the mice were sacrificed and ankles (A) and hearts (B) were processed for histology and scored for lesion severity. Levels of B. burgdorferi in tissue were determined by qRT-PCR (C). Data are representative of two separate experiments. n = 8, bars represent means ± SD. * vs. TLR2^−/− control, ** vs. WT CD4 or CD8-depleted mice, *** vs. WT control, p > 0.05.

Figure 6

Borrelia-specific antibody production. Mice were infected and treated as in Figure 4. On day 21 postinfection, the mice were sacrificed and assessed for B. burgdorferi-specific IgM (A) and IgG (B) levels in serum. Data are representative of two separate experiments. n = 3, bars represent means ± SD. *p > 0.05 compared to mouse strain control.

Figure 7

Neutrophil and macrophage numbers in infected ankle joints. Mice were infected and treated as in Figure 4. On day 21 postinfection, the mice were sacrificed and neutrophil (A) and macrophage (B) numbers were assessed from joint tissue. Data are representative of two separate experiments. n = 3, bars represent means ± SD. *p > 0.05.

Figure 8

Cytokine production in joints of infected mice. Mice were infected and treated as in Figure 4. On day 21 postinfection, the mice were sacrificed and levels of IFNγ (A), CXCL9 (B), and CXCL10 (C) were determined from homogenized joint tissue using a Lumina assay. CXCL9 levels within ICAM-1⁺VCAM-1⁺CD14⁻ synoviocytes were assessed from TLR2^−/− mice at 14 days postinfection by flow cytometry (D). n = 3, data representative of one trial. Bars represent means ± SD. *p > 0.01 compared to WT control.