Phospholipid-binding proteins differ in their capacity to induce autoantibodies and murine systemic lupus erythematosus

We have previously shown that immunization of nonautoimmune mice with the phospholipid-binding protein beta 2-glycoprotein I (beta 2GPI), in combination with lipopolysaccharide (LPS), induces a murine model of systemic lupus erythematosus (SLE), with sequential emergence of autoantibodies and glomerulonephritis. Here, we determine whether the paradigm for induction of murine SLE extends to other phospholipid-binding proteins. Mice were immunized with a phospholipid-binding protein (prothrombin (PT), protein S, or beta 2GPI), or a nonphospholipid-binding protein (glu-plasminogen), in the presence of LPS. The breadth and degree of the autoantibody response, and the frequency of glomerulonephritis, varied among the three proteins, with beta 2GPI being the most effective in inducing SLE-like disease. The phospholipid-binding proteins also differed in the pattern of serum cytokines they elicited. The most apparent difference between beta 2GPI and the other phospholipid-binding proteins was in their ability to bind to LPS: beta 2GPI bound to LPS, while PT and protein S did not. Our data suggest that binding to phospholipid(s) is a necessary, but not sufficient, condition for full induction of murine SLE. We propose that other properties, such as physiologic function, avidity for anionic phospholipids, and degree of interaction with other cell surface and/or circulating molecules (particularly LPS) may determine the range and severity of disease.