Validation of ERCC1-XPF Immunodetection – Response

Immunodetection and quantification of ERCC1-XPF in tumors may well develop into a clinically relevant technique, but it requires investigation with an antibody specific for the nuclease. Our article emphasizes that the reliability of immunohistochemistry depends upon rigorous selection of a specific antibody (1). As we pointed out, the antibody 8F1 does recognize ERCC1 without complications in HeLa cells, and this is apparently also the case for some other cancer cell lines as noted by Olaussen and Soria (in their Fig. 1A and B). Nevertheless, the existence of a strong cross-reacting protein in four human fibroblast cell lines raises great concern about the specificity of 8F1 and the validity of using it to draw conclusions about ERCC1 expression in tissue samples. The antibody FL297 is highly specific for ERCC1, if imperfect for immunohistochemistry. FL297 seems to stain nuclei in the control sample and the cytoplasm in the ERCC1 knockdown sample of Olaussen and Soria (their Fig. 1C), the same staining pattern we reported in Fig. 5 of our publication using genetic mutants (1). In contrast, 8F1 gives a varying degree of nuclear staining in all samples. We interpret the differential staining of normal and ERCC1-depleted cells with FL297 as an opportunity to test whether immunodetection of ERCC1 is clinically relevant. Readers are encouraged to examine the evidence presented in our article for the cross-reactivity of 8F1. It will indeed be of interest to identify the cross-reacting protein detected by 8F1. In the meantime, we and others are developing new antibodies and subjecting them to thorough quality control. Disclosure of Potential Conflicts of Interest R.D. Wood: coinventor with Cancer Research UK of antibody 8F1; advisory board, DNAR. The other authors disclosed no potential conflicts of interest.