Absence of Runx3 expression in normal gastrointestinal epithelium calls into question its tumour suppressor function

The Runx3 transcription factor regulates cell fate decisions during embryonic development and in adults. It was previously reported that Runx3 is strongly expressed in embryonic and adult gastrointestinal tract (GIT) epithelium (Ep) and that its loss causes gastric cancer. More than 280 publications have based their research on these findings and concluded that Runx3 is indeed a tumour suppressor (TS). In stark contrast, using various measures, we found that Runx3 expression is undetectable in GIT Ep. Employing a variety of biochemical and genetic techniques, including analysis of Runx3-GFP and R26LacZ/Runx3Cre or R26tdTomato/Runx3Cre reporter strains, we readily detected Runx3 in GIT-embedded leukocytes, dorsal root ganglia, skeletal elements and hair follicles. However, none of these approaches revealed detectable Runx3 levels in GIT Ep. Moreover, our analysis of the original Runx3LacZ/LacZ mice used in the previously reported study failed to reproduce the GIT expression of Runx3. The lack of evidence for Runx3 expression in normal GIT Ep creates a serious challenge to the published data and undermines the notion that Runx3 is a TS involved in cancer pathogenesis.

In an attempt to clarify the immunohistochemical basis for the conflicting results regarding the expression of Runx3 in GIT epithelium, we exchanged anti-Runx3 Abs with Yoshiaki Ito. Our lab provided the poly-G anti-Runx3 Abs (Fig 1 and 2) to Ito's lab and obtained three anti-Runx3 Abs designated R3-8C9, R3-3F12 and R3-1E10 (Ito et al, 2009). Mapping of the epitopes recognized by these Abs as reported by Ito K. et al ) is shown herein in Fig S1A. R3-8C9 recognizes an epitope at the C-terminal end of the protein, whereas the epitope recognized by R3-3F12 and R3-1E10 maps immediately downstream of the RUNT domain (Fig S1A), in a region defined by Ito Y. et al as GIT epithelium "exposed" ). An additional GIT epithelium "exposed" region , is located at the N-terminus of Runx3 ( Fig   S1A). This region coincides with the region recognized by the two anti-Runx3 Abs, Pep-J and GS, which readily detected Runx3 in GIT leukocytes, but failed to do so in the GIT epithelium ( Fig 2A&D).
Using the Poly-G Abs, which we provided, ) confirmed our previous observations (Levanon et al, 2001;Levanon & Groner, 2004); and data presented in the manuscript Fig 1 and 2), that Poly-G detect Runx3 in DRG, but not in GIT epithelium ). Using R3-3F12 and R3-8C9 Abs in IHC of GIT sections we found that the GIT leukocytes in each section were clearly stained, whereas the GIT epithelium was not ( Fig S1B). Similar results were obtained with Poly-G ( Fig S1B). Of note, to avoid the high background staining, that occurs when using mouse primary antibodies on mouse tissues, the analysis presented in Fig   S1B&C, was conducted with the MOM kit (Vector laboratories Burlingame, CA, USA), designed specifically to use with mouse primary Abs on murine tissues. This protocol was particularly important because GIT epithelial cells are notorious for their high degree of nonspecific antibody binding.
It was previously reported by Ito K. et al ) that of the three monoclonal anti-Runx3 Abs (R3-1E10, R3-3F12 and R3-8C9) R3-1E10 displayed an inverse reactivity pattern compared to Poly-G, namely, it detected Runx3 in GIT epithelium, but not in the DRG ). This was a peculiar finding since Runx3 expression in DRG is considered the undisputed golden standard of the Runx3 expression-signature (Inoue et al, 2002;Kramer et al, 2006;Levanon et al, 2002;Levanon et al, 2001;Li et al, 2002;Marmigere & Ernfors, 2007). To explain the unusual activity of R3-1E10 Abs the authors postulated that in GIT epithelium the Cterminal region of Runx3 is sequestered in an epithelial-specific manner . Two uncommon mechanisms were suggested. The first postulated that the GIT epithelial C-terminal region of Runx3 undergoes a conformational change due to an epithelial-specific modification of the protein. The second suggested that an unidentified epithelial protein was specifically bound to the C-terminal region and masked certain epitopes . Based on these assumptions, neither of which were supported by any experimental evidence, the authors concluded that the reason Poly-G failed to detect Runx3 in GIT epithelium is because it was raised against the C-terminal region of Runx3, which is inaccessible in epithelial cells ).
Using R3-1E10 Abs on sections of either DRG or GIT tissues, we confirmed the finding of Ito et al ) that R3-1E10 Abs did not detect Runx3 in DRG ( Fig S1C), but we further found that R3-1E10 neither reacted with GIT epithelial cells nor with GIT-embedded leukocytes ( Fig S1C). Of note, the inability of R3-1E10 to detect Runx3 in DRG and/or in leukocytes, the two major and undisputed sites of Runx3 expression, disqualifies it as a valid anti-Runx3 Ab. Supporting this conclusion are the findings that Mono-G Abs, which were raised against the same GIT epithelium "exposed" region ( Fig S1A) readily reacted with Runx3 in DRG, but not in GIT epithelium ( Fig S1C). Hence, results attained using R3-1E10 should be interpreted cautiously, in particular its reaction with GIT epithelium ).
Together, the combined outcome of the IHC experiments (Fig S1B-C) and the comprehensive evidences for the absence of Runx3 in GIT epithelium described in the manuscript, exclude the possibility of a GIT epithelium-specific sequestering of Runx3 (Ito et al, 2009; and demonstrate that R3-1E10 is not a valid anti-Runx3 Ab. In all, more than seven anti-Runx3 Abs were tested, either in the present manuscript or by Ito et al ; of these, only R3-1E10 did not detect Runx3 in DRG and leukocytes. Therefore, as noted before, it is unfortunate and scientifically unsound that of all available anti-Runx3 Abs, R3-1E10 was the one recently used to detect Runx3 expression in GIT epithelium ) without a proper disclosure of any information regarding its unique properties. Given that GIT epithelial cells are notorious for their high degree of nonspecific antibody binding, it is important to note that of the seven anti-Runx3 Abs raised by different laboratories not a single one produced a reliable signal when reacted with GIT epithelium (Fig 2   and Fig S1). This finding poses a serious challenge to published data documenting Runx3 expression in GIT epithelium that was solely based on IHC. The data presented herein which using various measures demonstrate that Runx3 expression in undetectable in GIT epithelium strongly support this conclusion.  detected Runx3 in the stomach epithelium. Of note, these two monoclonal anti-Runx3

Supporting Information References
Abs (Mono-G and R3-1E10) were raised against similar Runx3 region (A). In E14.5 DRG (bottom panels) Runx3, which is highly expressed in TrkC neurons (Inoue et al, 2002;Kramer et al, 2006;Levanon et al, 2002;Levanon et al, 2001;Li et al, 2002;Marmigere & Ernfors, 2007), is readily detected by Mono-G, but not by R3-1E10. A. Scheme of RUNX3 protein structure indicating the position of the peptide used for generation of the Active Motif (AM) anti RUNX3 antibody. The position of Poly-G is shown in Fig S1. B. Human cardiopyloric stomach sections (Upper panels x4, lower panels x40) were reacted with pre-immune serum (left panels), with AM Ab (middle panels), or with Poly-G anti RUNX3/Runx3 Ab. The pre-immune serum as well as the AM and Poly-G Abs reveal the characteristic high background staining in the epithelium. Specific staining of leukocytes is detected in the AM and Poly-G but not the pre-immune panels. Table 1 Table S1. List of publications addressing associations of Runx3 loss with tumor development in wide spectrum of cancers

Supporting Information
The list was compiled from PubMed using "Runx3" and "Cancer" as search terms.  (2007) Hypermethylation of the COX-2 gene is a potential prognostic marker for cervical cancer.