I recall in some papers the testing of dimerized beta2-GPI. It does bind to PS, but then it doesn't have the Fc end of the antibody and so macrophages, DCs can not bind to it. When experiments are done with bavi in mice human beta2-GPI is also injected because bavi binds to the second domain of human beta2-GPI, and not to the first domain which is conserved across many species. I posted this once.