However, the percentage of patients and controls expressing these antibodies show large variations between studies (Nakamura et al., 1998, Treon et al., 2000, von Mensdorff-Pouilly et al., 2000a, von Mensdorff-Pouilly et al., 2000b and Apostolopoulos et al., 2006). In some studies, MUC1 serum antibodies could not be detected in healthy controls (Apostolopoulos et al., Regorafenib clinical trial 2006), whereas other studies demonstrated that up to 16% of healthy controls show reactivity to MUC1 peptides (Nakamura et al., 1998).

In cancer patients, the reported levels of anti-MUC1 antibodies also differ, due to the presence of soluble serum MUC1. Depending on tumour type, these serum MUC1 antigens have been shown to complex with anti-MUC1 antibodies (Treon et al., 2000). Standardization of the different methods, including the flowcytometric assay we describe, seems to be necessary to answer the question on prevalence of anti-MUC serum antibodies in healthy controls and cancer patients. Smad inhibitor The numbers of samples tested in this study does not justify a conclusion on prevalence of these antibodies; we merely show that with this technique we are able to detect human serum antibodies directed to MUC1 and underglycosylated MUC1. In addition to the detection of

serum antibodies against unglycosylated MUC1, manipulation of MUC1 glycosylation in the CHO-ldlD MUC1 system allowed us to selectively test for the presence of IgG and IgM antibody responses to MUC1-Tn. These serum antibodies could only be detected in a breast cancer patient after vaccination and not in non-vaccinated cancer patients or healthy controls. Detection of antibodies directed to underglycosylated MUC1 has been recently described by Wandall et al. (2010), who made use of an O-glycopeptide microarray to demonstrate

that MUC1-Tn/STn associated IgG serum antibodies are present in low numbers of newly diagnosed breast, ovarian and prostate Unoprostone cancer patients and not in healthy controls. Additionally, in patients who had no pre-existing MUC1-Tn/STn IgG antibodies, it was shown that they did develop detectable serum IgG and IgM MUC1-Tn antibodies after vaccination. Similar findings were previously described by Sabbatini et al. (2007), who demonstrated that MUC1-Tn antibodies could be detected by ELISA. Both ELISA and O-glycopeptide microarrays make use of small MUC1 peptides that are differently glycosylated. The O-glycopeptide microarray allows rapid mapping of serum antibody specificity and has already been proven to be reliable in detection of MUC1 serum antibodies in mice vaccination studies ( Westerlind et al., 2009). Even though the glycosylation sites can be controlled in the small peptide-based methods, allowing specific antibody mapping, these methods are only able to detect antibodies binding to linear MUC1 structures.