COMMENTARY plant immunologists to correct their original mistakes, as transplants engineered to express FasL could prove to be disastrous, leading to rapid rejection and abscess formation rather than increased tissue engraftment. Vaux, who wrote one of the earliest and clearest descriptions of FasL as the ‘enforcer’of immune privilege, took the unusual step of retracting his News and Views piece after his own lab found inflammation, not immunosuppression2. The continued proliferation of erroneous ideas about FasL may have been due in part to technical problems hindering early research. A principal problem involved the antibodies used in the early reports. A polyclonal antibody against human FasL (C-20), produced by Santa Cruz Biotechnologies, was not highly specific22, leading to the publication of false-positive results on a least a half-dozen different occasions by the time its cross-reactivity was discovered in 1998, and in several reports by others since then23. Compounding the problem may be the fact that many other polysera were generated in a similar way24, but were not tested in 1998 report22. An underlying and ongoing problem is the use of polysera made by injecting animals with peptides. Although the generation of polyclonal antibodies is straightforward, a complete characterization of antibody specificity is difficult and fraught with pitfalls. Another widely used reagent, a mouse monoclonal antibody (clone 33) from Transduction Laboratories, was also reported to be nonspecific for FasL (ref. 25). Other possible confounding variables include the use of nonintron-spanning PCR primers without proper controls7, 26, contamination of fresh tumor samples with lymphocytes (which can express large amounts of FasL) and problems with functional assays. For example, several studies involved T-cell targets that were themselves capable of expressing FasL, such as Jurkat cells. When target cell death was found and blocked with antibodies against FasL, researchers were not always able to verify that the cell death was induced by FasL expression on tumor cells and not by induced expression of FasL on T cells26. Thus, the devil really was in the details.

One consequence of these misunderstandings, for tumor immunology, was that researchers have been side-tracked in determining the true biological function of Fas and FasL as mediators of cytotoxicity27 and as central mediators of activation-induced cell death28. Thus, it is the case that Fas-FasL interactions can cause T cell death and this death is important in the induction of tolerance, immune homeostasis and lymphocyte effector functions. Indeed, it has been confirmed that melanoma-specific T lymphocytes undergo apoptotic death after the major-histocompatibility-complex-restricted recognition of tumor cells, and T-cell death can be blocked by the addition of a specific antibodies against Fas (ref. 29). However, contrary to the prevailing view that tumor cells cause the death of anti-tumor T cells by expressing FasL, it is now apparent that in most cases, FasL is expressed by T lymphocytes upon activation after tumor cell recognition, causing them to kill themselves (‘suicide’) and each other (‘fratricide’)(Fig. 1) 16, 29, 30. When FasL is expressed ectopically with the goal of inducing the death of T lymphocytes, researchers must consider the resultant caspase cascade and the potential for the consequent activation of an innate immune response. Hindsight is always perfect, and it is now possible to view initial mistakes and contradictory data regarding the function of FasL in a new light. A tantalizing new idea, featured prominently in the literature, can rapidly spread through the scientific …