Effect of removing human Graves' thyroid xenografts after eight weeks in nude mice and rexenografting them into SCID mice

Human thyroid xenografts from four patients with Graves' disease (GD) and two normal persons were initially xenografted into nude mice. Eight weeks after xenografting, the thyroid tissue appeared normal; indeed, thyroid infiltrating lymphocytes in the GD xenograft could no longer be identified when analyzed histologically. Thus, human immunoglobulin G (IgG), thyroperoxidase (TPO)-antibodies (Abs), thyroglobulin (Tg)-Abs, thyroid- stimulating antibodies (TSAb), and thyrocyte histocompatibility leucocyte antigen (HLA)-DR expression were undetectable. These same tissues were retrieved from the nude mouse and rexenografted into severe combined immunodeficient (SCID) mice (with no prior xenograft); autologous peripheral blood mononuclear cells (PBMC) or CD8-depleted PBMC (non-CD8 cells) were simultaneously injected into some of these SCID mice. Engraftment of a GD thyroid rexenograft (TH) alone did not cause IgG, TSAb, TPO-Ab, or Tg-Ab production, thyrocyte HLA-DR expression, or lymphocytic infiltration in thyroid grafts. Engraftment of GD PBMC or non-CD8 cells alone (i.e. without a thyroid xenograft) caused human IgG to rise, but only minimal titers of thyroid antibodies appeared. When TSAb, TPO-Ab, and Tg-Ab were quantified, GD TH plus PBMC-engrafted SCID mice showed significantly higher production of each antibody than that of GD PBMC alone, and this phenomenon was further enhanced by the removal of CD8⁺ cells. GD thyrocytes showed marked HLA-DR expression at human surgery; however, after 8 weeks' sojourn in nude mice, DR expression disappeared. After a further 8 weeks following rexenografting into SCID mice, TH plus PBMC resulted in a reappearance of DR expression only in GD but not in grafts from normal persons, and this was enhanced by the depletion of CD8 cells. These results were also in parallel with histological findings inasmuch as the normal tissue remained normal with no thyroid antibodies appearing with PBMC or CD8-depleted cells. In experiments from two GD patients, autologous skeletal muscle as well as thyroid tissue were xenografted into nude mice. Eight weeks after xenografting, these were rexenografted into SCID mice that contained prior autologous primary GD thyroid xenografts. Histological findings showed new lymphocytic infiltration in rexenografted thyroid tissues in the SCID mice but not in autologous skeletal muscle. This signifies that the immune assault in GD is specifically targeted to the thyroid tissue. In conclusion, 1) intrathyroidal lymphocytes from primary GD thyroid xenografts in SCID mice migrated to the autologous second thyroid xenografts (rexenografted from nude mice) but not to autologous skeletal muscle; 2) engraftment of GD PBMC or non-CD8 cells alone caused human IgG to rise, but thyroid antibodies either did not appear or were present in low titer; and 3) rexenografted GD thyroid tissue from nude to virgin SCID mice plus autologous PBMC engraftment caused thyroid antibody production and thyrocyte DR expression to rise. This was further enhanced by the depletion of CD8⁺ cells. This animal model may help to elucidate the pathogenesis of human autoimmune thyroid disease.