Breaking tolerance in transgenic mice expressing the human TSH receptor A-subunit: thyroiditis, epitope spreading and adjuvant as a 'double edged sword'

Abstract

Transgenic mice with the human thyrotropin-receptor (TSHR) A-subunit targeted to the thyroid are tolerant of the transgene. In transgenics that express low A-subunit levels (Lo-expressors), regulatory T cell (Treg) depletion using anti-CD25 before immunization with adenovirus encoding the A-subunit (A-sub-Ad) breaks tolerance, inducing extensive thyroid lymphocytic infiltration, thyroid damage and antibody spreading to other thyroid proteins. In contrast, no thyroiditis develops in Hi-expressor transgenics or wild-type mice. Our present goal was to determine if thyroiditis could be induced in Hi-expressor transgenics using a more potent immunization protocol: Treg depletion, priming with Complete Freund's Adjuvant (CFA) + A-subunit protein and further Treg depletions before two boosts with A-sub-Ad. As controls, anti-CD25 treated Hi- and Lo-expressors and wild-type mice were primed with CFA+ mouse thyroglobulin (Tg) or CFA alone before A-sub-Ad boosting. Thyroiditis developed after CFA+A-subunit protein or Tg and A-sub-Ad boosting in Lo-expressor transgenics but Hi- expressors (and wild-type mice) were resistant to thyroiditis induction. Importantly, in Lo-expressors, thyroiditis was associated with the development of antibodies to the mouse TSHR downstream of the A-subunit. Unexpectedly, we observed that the effect of bacterial products on the immune system is a "double-edged sword". On the one hand, priming with CFA (mycobacteria emulsified in oil) plus A-subunit protein broke tolerance to the A-subunit in Hi-expressor transgenics leading to high TSHR antibody levels. On the other hand, prior treatment with CFA in the absence of A-subunit protein inhibited responses to subsequent immunization with A-sub-Ad. Consequently, adjuvant activity arising in vivo after bacterial infections combined with a protein autoantigen can break self-tolerance but in the absence of the autoantigen, adjuvant activity can inhibit the induction of immunity to autoantigens (like the TSHR) displaying strong self-tolerance.

Figure 1. Protocol for breaking tolerance to the human TSHR in Hi-expressor A-subunit transgenics by priming with one injection of CFA+TSHR A-subunit protein CFA followed by boosting twice with A-subunit-adenovirus.

Parallel studies were performed in Lo-expressor transgenics and wild-type littermates. Four days before each immunization, mice were injected with anti-CD25 to deplete regulatory T cells (Treg). The same protocol was used in a second set of mice to test the outcome of immunization with CFA+mouse Tg protein or, in a third set, with CFA alone.

Figure 2. Antibodies to the TSHR and Tg induced in transgenic mice expressing high or low levels of the human TSHR A-subunit (Hi - and Lo- expressors, respectively) and wild-type littermates.

Mice were depleted of CD25-positive cells before priming with (a) CFA+TSHR A-subunit protein; (b) CFA+mouse Tg or (c) CFA alone. Subsequently, mice received two cycles of anti-CD25+ depletion 4 days before immunization with A-sub-Ad. Data shown are the Mean+SEM (n = 5–7 mice) in sera drawn after 32 days (striped bars) and after 74 days (black bars; euthanasia). TSHR protein and TSHR antibodies (ELISA and TBI) are indicated in red; mouse Tg and mouse Tg antibodies are indicated in green. The grey areas indicate the mean ± 2 SEM for wild-type mice immunized with control-adenovirus (n = 4). Statistical analyses:- Upper panel-TSHR ELISA antibody: Significantly greater in WT than Hi-expressor mice at day 74 * p = 0.004; ^#, p = 0.016 (rank sum test). Middle panel - TBI: significantly greater in WT mice after 65 days than in Hi-expressors at day 65 # p<0.05 (ANOVA), * p<0.001 (t test); ∧ significantly lower than in WT mice immunized three times with A-sub-Ad p<0.001 (t test). Lower panel - Tg antibody: *, p<0.05 versus Hi-expressor and WT mice (ANOVA, p<0.5).

Figure 3. Antibodies to TPO, thyroiditis and serum T4 in Hi-expressor transgenics, Lo-expressor transgenics and WT mice immunized with CFA+TSHR A-subunit protein, CFA+mouse Tg, or CFA alone, followed by two immunizations with A-sub-Ad (see legend to Fig. 2 ).

Four days prior to each of the three immunizations, mice were treated with anti-CD25 to deplete regulatory T cells. Data shown are the Mean+SEM (n = 5–7 mice) in sera drawn after 32 days (striped bars) and 74 days (euthanasia; black bars). The grey areas indicate the mean ± 2 SEM for wild-type mice immunized with control-adenovirus (n = 4). Statistical analysis:- Middle panel- Thyroiditis: * significantly greater in Lo expressors primed with CFA+A-sub, Tg or PBS than in Hi-expressor transgenics or wild-type mice (ANOVA on ranks, p<0.05); #, significantly greater in Hi-expressors primed with CFA+Tg than in the same strain primed with CFA+A-sub or CFA+PBS (ANOVA on ranks, p<0.05). Lower panel: * Serum T4 is significantly lower in Lo-expressor transgenics immunized with CFA+A-sub or CFA+Tg than in Hi-expressor transgenics or wild-type mice (p<0 05, ANOVA).

Figure 4. IgG subclasses of TSHR antibodies in Hi- and Lo expressor A-subunit transgenic mice and WT littermates injected with anti-CD25 before immunization with TSHR A-sub protein+CFA and two subsequent injections of human TSHR A-subunit-adenovirus.

ELISA plates were coated with TSHR A-subunit protein and serum binding was detected using goat anti-mouse IgG1, IgG2a and IgG2b. Data are shown for the mean+SEM OD 490 nm in ELISA for 4 WT mice, 7 Lo-expressor transgenics and 5 Hi-expressor transgenics.

Figure 5. Linear TSHR antibody epitopes recognized by Hi- and Lo expressor A-subunit transgenic mice and wild-type littermates injected with anti-CD25 before immunization with TSHR A-sub protein+CFA and two subsequent injections of human TSHR A-subunit-adenovirus.

Antibody binding was studied to ELISA wells coated with synthetic peptides encompassing the human TSHR ectodomain (peptides A to Z) and two extracellular loops (EC1 and EC2)(amino acid sequences in Table 1). A dashed line indicates the last peptide (“R”) in the human TSHR A-subunit. The shaded area (to the right) includes the human peptides downstream of the A-subunit (S to Z, EC1 and E2); antibody binding to peptides in this region implies recognition of the mouse TSHR. The data for individual mice are shown as the OD binding value (490 nm) for each peptide (black circles) and to TSHR A-sub protein (black bar). The number of mice studied in each group is given in parentheses.