Cell-Mediated Cytotoxicity in Lyme Arthritis

Abstract

Objective: Obliterative microvascular lesions are found in the synovial tissue of ~50% of patients with post-antibiotic Lyme arthritis (LA) and correlate with autoantibodies to certain vascular antigens. In this study, we identified lymphocytes with cytotoxic potential that may also mediate this feature of synovial pathology.

Methods: The cytotoxic potential of lymphocytes and their T cell receptor (TCR) V_β gene usage were determined using samples of peripheral blood mononuclear cells (PBMCs) and synovial fluid mononuclear cells (SFMCs) from patients with antibiotic-responsive or post-antibiotic LA. Cell phenotypes were analyzed using flow cytometry and intracellular cytokine staining. Immunohistochemistry was performed on post-antibiotic synovial tissue samples.

Results: In SFMC and PBMC samples, the percentages of CD8+ T cells and double-negative T cells (primarily γδ T cells) were greater among 22 patients with post-antibiotic LA than in 14 patients with antibiotic-responsive LA. Moreover, CD8+ T cells and γδ T cells often expressed cytotoxic mediators, granzyme A/granzyme B, and perforin. The same 3 TCR V_β segments were over-represented in both CD4+ T cells and CD8+ T cells in SFMC samples from post-antibiotic LA patients. In synovial tissue samples from 3 patients with post-antibiotic LA, CD8+ T cells intermixed with CD4+ T cells were seen around blood vessels, and 2 patients with microvascular damage had autoantibodies to vascular-associated antigens. One of these 2 patients, the one in whom cytotoxicity appeared to be active, had complement (C5b-9) deposition on obliterated vessels. Very few natural killer cells or γδ T cells were seen.

Conclusion: We propose that CD8+ T cell-mediated cytotoxicity, CD4+ T cell help, autoantibodies to vascular antigens, and complement deposition may each have a role in microvasculature damage in post-antibiotic LA.

Figure 1.. Lymphocyte frequencies in PBMC or SFMC from LA or RA patients and in healthy subjects.

In panel A, the gating strategy to define CD4, CD8, and double-negative T cells and NK cells is shown. In panel B, lymphocyte frequencies are compared for each cell population in healthy control subjects (HC, n=15), and in patients with antibiotic-responsive LA (Resp LA, n=15), post-antibiotic LA (Post LA, n=22) or rheumatoid arthritis (RA, n=7). In panel C, the gating strategy is shown for cytotoxic CD4+CD28−SLAMF7+ T cells in PBMC and SFMC of one LA patient, which shows that these cells did not express CD28 and were not found in SFMC. In panel D, the frequency of CD4+CD28−SLAMF7+ T cells within total CD3+ lymphocytes is shown in PBMC for each patient group. Differences between groups were analyzed by unpaired t-test with Welch's correction. The comparisons achieving statistical significance are shown as follows: * = P≤0.05, **P≤0.01, ***P≤0.001.

Figure 2.. Double negative T cell subpopulations in PBMC or SFMC in LA patients.

In panel A, the 3 types of CD4 and CD8 double-negative T lymphocytes are represented. Below, the gating strategy is shown for the identification of the 2 major γδ T cell populations (Vδ1 and Vδ2 T cells), and for αβ TCR T cells; iNKT cells were determined by Vβ11 expression (PE, reaction G in IOTest^® Beta Mark, Beckman Coulter). In panel B, the percentages of Vδ1 and Vδ2 γδ T cells, αβ TCR T cells and iNKT cells within the total CD3+ T cell subset are compared in patients with antibiotic-responsive LA (Resp LA, n=8) or post-antibiotic LA (Post LA, n=12). Differences between groups were analyzed by unpaired t-test with Welch's correction. The comparisons achieving statistical significance are shown as follows: * = P≤0.05.

Figure 3.. Intracellular staining of CD4+, CD8+ and DN T cells and NK cells in SFMC and synovial fluid cytotoxic effector proteins in LA and RA patients.

The percentage of cells expressing cytotoxic mediators (granzyme A, granzyme B, perforin 1, or the proinflammatory mediator interferon gamma [IFNγ]) are shown in patients with antibiotic-responsive LA (Resp LA, n=14) or post-antibiotic LA (Post LA, n=19) or rheumatoid arthritis (RA, n=5). The gating strategy for the cell populations is shown in figure 1A. Differences between groups were analyzed by two-tailed unpaired t-test with Welch's correction. The comparisons achieving statistical significance are shown as follows: * = P≤0.05, **P≤0.01, ***P≤0.001.

Figure 4.. Identification of Vβ gene usage of CD4+, CD8+, and DN T cell populations in SFMC in LA patients.

T cell receptor V beta (Vβ) gene usage overrepresented in CD4+, CD8+, and DN T cells from inflamed joints were determined by comparing ratios of Vβ usage in SFMC vs. patient-matched and date-matched PBMC from patients with antibiotic-responsive LA (n=5) and post-antibiotic LA (n=5). The dotted line indicates a SFMC/PBMC ratio of 1. Statistically significant differences in SFMC/PBMC ratios of Vβ usage between patient groups were determined by two-tailed unpaired t-test with Welch’s correction. The comparisons achieving statistical significance (* = P≤0.05, ** = P≤0.01) are underlined.

Figure 5.. Immunohistologic staining pertaining to microvascular damage in synovial tissue from 3 patients with post-antibiotic LA.

In the 3 patients, synovial villi are shown with staining for CD31 (endothelial cells), CD8+ T cells, CD4+ T cells, C5b-9 (the complement attack complex), and for NKG2D, an activating receptor. In patient #2, microvascular damage appears active. In this patient, CD8+ T cells are seen in angiocentric locations; CD4+ T cells are spread diffusely in sublining areas and intermixed with CD8+ T cells, but CD4+ cells are not in contact with endothelial cells; C5b-9 staining is found in partially obliterated or completely obliterated (inset) blood vessels; and positive staining for NKG2D receptors is also seen in angiocentric locations. Patient #1 has patent blood vessels and lacks C5b-9 or NGK2D staining, whereas patient #3 has complete obliteration of almost all blood vessels, but the immune process against obliterated vessels appears inactive. In both patients 1 and 3, non-specific background fluorescent staining is seen on blood vessels stained for complement components C5b-9. Pictures are shown at 20 X 100 magnification except for CD4+T cells in patient 2, which are shown at both low power (10 X 100) and high power (inset).

Figure 6.. Autoantibody responses to 3 host vascular antigens in 3 patients with post-antibiotic LA.

Serologic responses to 3 vascular-associated autoantigens, endothelial cell growth factor (ECGF), annexin A2 (annexin), and apolipoprotein B-100 (apoB-100), as determined by ELISA, are shown in the 3 patients who underwent synovectomies for the treatment of post-antibiotic LA in whom immunohistologic staining of synovial tissue is shown in figure 5. Patients 2 and 3 had obliterative microvascular lesions in synovial tissue, and they both had positive IgG antibody responses to 2 or 3 autoantigens, respectively. Patient 1 did not have antibody responses to these autoantigens, and obliterative microvascular lesions were not seen in his synovial sample. Absorbance values >2SD above the mean values of normal controls were defined as positive. For ECGF, the cut-off value was 0.42; for annexin A2, is was 0.35; and for apoB-100, it was 0.22. Detailed methods for ELISA assays are given in references -.