Adenocarcinoma of the colorectum (CRC) is the third leading cause of cancer-related death in the United States as well as worldwide (1). This is due in large part to discovery at advanced stages of disease, with only 39% of patients being diagnosed at a localized disease stage (2). Histopathologic evaluation of endoscopically sampled colonic lesions has thus come to play a critical role in the accurate and expedient diagnosis of CRC. CRC develops through the progressive transformation of the colonic mucosa from normal to dysplastic precursor lesion and ultimately to malignancy. The morphologic and molecular genetic stages in this progression are increasingly well characterized (3, 4). Alterations in glycan processing of membrane glycoproteins in CRC have been appreciated for some time (5–7), but the exact sequence of glycan changes, and relationship to early stages in carcinogenesis, are less well understood.
Under physiologic circumstances, O-linked glycosylation is a universal post-translational modification, and abnormalities in O-glycosylation have been described in many carcinomas, including CRC (7, 8). Among the O-glycan changes associated with CRC is the expression of specific, abnormally truncated glycans (9), such as the Tn neoantigen (CD175, GalNAcα1-O-Ser/Thr) and its sialylated form sialyl Tn (STn) (CD175s, Neu5Acα2-6GalNAcα1-O-Ser/Thr). These are tumor-associated carbohydrate antigens (TACAs) which have been identified in a broad spectrum of adenocarcinomas including CRC (10, 11).
The Tn antigen is a precursor structure biosynthesized in the Golgi apparatus by a family of twenty different polypeptide-N-acetylgalactosaminyltransferases (ppGalNAc-Ts), which transfer GalNAc from the donor UDP-GalNAc to a Ser or Thr residue in glycoproteins. The Tn antigen subsequently serves as the common core to which glycans are extended first by the T-synthase and its private molecular chaperone Cosmc to form primarily core 1 structures (Galβ1-3GalNAcα1-O-Ser/Thr, the T or TF antigen), which are precursors to branched core 2 structures; other O-glycans in colorectal glycoproteins include core 3 and core 4 O-glycans (12–14).
In most healthy tissues, the Tn antigen does not accumulate due to its complete conversion to extended O-glycans. In pathologic states including various malignancies, however, increased expression of the Tn and STn antigens has been correlated with tumor progression, metastasis, and poor prognosis (15).
Patterns of Tn antigen expression in CRC have been assessed historically using GalNAc-binding lectins, such as Vicia villosa agglutinin (VVA) and Helix pomatia agglutinin (HPA), or with antibodies that are often inadequately characterized for specificity (12). These approaches have shown elevated but variable Tn expression in CRC, in up to 90% of tumors, with STn antigen expression paralleling that of the Tn antigen (16–18). However, very few studies have directly compared paired tumor and normal colorectal mucosa from the same individual. Instead, they have used samples from healthy subjects as control tissue and therefore lack internal controls (10).
Prior investigation of Tn expression in paired tumor and normal colorectal tissue samples has also been conducted using the anti-Tn monoclonal IgM BaGs6 (CA3638) which was obtained from the ascites of mice immunized with Tn-expressing cells. BaGs6 has been described as recognizing glycoconjugates containing GalNAcα1-O-Ser/Thr but not blood group A or similar glycans terminating in GalNAc, though its formation is experimentally variable due to the nature of the use of unmodified mouse ascites (19, 20). The Tn antigen has been detected using this reagent in over 90% of CRC, seldom in normal colorectal tissue, and to some degree in histologically normal-appearing peritumoral mucosa, though too few peritumoral samples have been examined to establish definitive Tn expression patterns (18).
We generated a recombinant antibody derived from BaGs6 and produced it in both a recombinant murine IgM form (ReBaGs6) and a human IgG1 form (Remab6) (21). ReBaGs6 is recombinant expressed in human HEK293 cells and provides a reagent with remarkably high affinity and specificity for clustered Tn antigen. This has been previously validated using a Tn glycopeptide microarray containing various peptides with one or more Tn antigens, including peptides modeled to mimic the hinge region of human IgA1; while ReBaGs6 demonstrated high affinity for di- and tri-Tn structures on mucin-derived glycopeptides, it only weakly bound to glycopeptides modeled after IgA1 and other non-Tn terminal GalNAc structures such as blood group A glycans (21). Additionally, ReBaGs6 has demonstrated efficacy in immunohistochemical applications. While binding of ReBaGs6 was not observed in normal murine intestinal tissues, extensive staining was observed in tissues from mice engineered to express Tn antigen via intraepithelial knockout of COSMC (C1GalT1C1), which encodes an essential chaperone required for elaboration of O-glycans under physiologic circumstances (21).
The objective of this study is therefore to provide the first detailed immunohistochemical analysis of Tn and STn expression patterns in CRC, and in matched benign peri-tumoral colonic mucosa, and mucosa distant from the tumor, using these well-defined monoclonal antibodies, with comparison to VVA lectin staining.