The specific detection of APC is of paramount importance due to its central role as a tumour suppressor protein, and its ability to result in the development of carcinomas when mutated. Here, we describe the detection specificities of a panel of antibodies that have been developed to detect APC. This work extends several recent publications in this area and importantly serves to provide a comprehensive analysis of APC antibody specificities (Roberts et al, 2003
; Brocardo et al, 2005
Initially, we carried out Western blotting experiments, using a panel of antibodies to APC to confirm the size of expressed APC in a number of different cell lines. We observed a band at approximately 150
kDa with a number of antibodies in the three different cell lines used. One possible explanation for the presence of this band was expression of an unusual and cryptic splicing event in exon 15 of APC. This 150-kDa protein could not have been a post-translational modification, as it was detected by C-terminal-directed antibodies to APC, beyond the termination codon in SW480, in addition to it being detected with N-terminal antibodies. We then carried out further experiments to confirm or refute this based on validating the specificities of the antibodies used to detect APC, as an alternative explanation for the presence of this 150-kDa band on APC Western blots could be that several antibodies were crossreacting with proteins of approximately the same size.
Two approaches were followed to confirm the specificity of antibodies through (i) reciprocal co-IP and blotting experiments and (ii) RNAi experiments. The former approach suggested the presence of a 150-kDa cryptic splice variant of APC, or that several APC antibodies were all crossreacting with the same 150-kDa protein. The latter approach suggested that the antibodies used were likely to be detecting a protein other than APC at 150
kDa, as this 150-kDa protein could not be knocked down by APC RNAi. The apical protein seen on IF staining with several APC antibodies also failed to be knocked down by APC RNAi.
We considered the possibility that a cryptic splicing event produces a variant that is more stable than ‘conventional' APC. The region of APC between amino acids 1034 and 1337 is responsible for the ubiquitin-proteasome-mediated downregulation of APC (Choi et al, 2004
). This region also encompasses the MCR and the truncating mutation in SW480. It remains a feasible proposition that a cryptic splicing event at the start of exon 15 may give rise to a form of APC that has this region spliced out, thereby, becoming more stable and therefore not ‘knocked down' under the conditions used here. However, RT–PCR with primers to exons 1,2 and 3 of APC and two primers within the sequence corresponding to the last 300 amino acids at the C terminus of APC, successfully amplified full-length APC, but did not give any products which would account for a spliced form of APC 150
kDa in size (data not shown). Northern blot gave a band at the expected size for full-length APC and no smaller bands (data not shown). As no alternatively spliced APC mRNA could be detected by either of these methods, it seems unlikely that the 150-kDa protein we observed could be an alternatively spliced form of APC, which is more stable than the full-length or truncated APC proteins expressed in the cell lines used in the RNAi experiments (Brocardo et al (2005)
looked for C-terminal splice variants of APC by Western blotting, but did not detect bands corresponding to potential splice variants. One antibody used in their studies, Ab4, is not recommended for use in Western blotting and fails to detect full-length APC (, ), the two other C-terminal antibodies used are not listed in our report).
Our findings with this panel of antibodies led us to further clarify and present a summary of the specificity of the panel of antibodies to APC that were used in this study. Many of these antibodies are able to reliably detect full-length APC and/or truncated APC. Of the antibodies tested by us, only Ab4 and N-15 failed to detect full-length or truncated APC. However, all the antibodies tested, which were able to detect APC, also gave several nonspecific bands. Five out of all the antibodies studied detect an apical protein by IF; APC RNAi confirmed that this protein is not likely to be APC and therefore it appears that these antibodies cannot reliably detect conventional APC by IF staining. Validation of the specificity of these antibodies will serve to lessen ambiguities in the literature.