Pseudomyxoma peritonei (PMP) is characterized by the growth of intestinal epithelial cells with extensive mucin secretion in the abdomen and pelvis. The cells grow freely in the peritoneal cavity and may be a low-grade mucinous neoplasm or a peritoneal mucinous carcinoma. However, the mucinous nature and the peritoneal location make it difficult to determine whether the neoplastic epithelial cells are benign or malignant. Consequently, it is difficult to determine how the tumor will behave, predict the long-term outcome, or plan rational therapy.
It has long been appreciated that mucins in the normal colon have a homogeneous pattern of terminal glycosylation that is altered in the mucins associated with colorectal cancer (CRC), which can be recognized by lectin binding patterns.1 The goal of this study was to use lectin binding patterns in the setting of PMP to characterize mucins secreted by low-grade neoplasms and those produced by adenocarcinomas using fluorescein isothiocyanate (FITC)-labeled lectins: FITC–Dolichos biflorus agglutinin (DBA), FITC–soybean agglutinin, FITC–Ricinus communis agglutinin-1, FITC–Ulex europeus agglutinin-1, and FITC–peanut agglutinin (PNA). We hypothesized that there would be differences in glycoconjugate structure that would differentiate mucins in adenocarcinomas from those in lower grade neoplasms.
The PMP cases were divided into 2 pathologic groups (low-grade benign mucinous neoplasms and mucinous adenocarcinomas), and the lectin binding was independently determined to be positive or negative for each case. Forty-five patients were studied; 25 had adenocarcinoma, and 20 a low-grade neoplasm (Table 1).
Table 1.
Demographics
| All | Type of neoplasm |
P value | ||
|---|---|---|---|---|
| Adenocarcinoma | Low-grade | |||
| Age, y, mean ± SD | 56.2 ± 16.1 | 56.3 ± 13.4 | 56.0 ± 19.4 | .94a |
| Male gender, n (%) | 19 (42.2) | 11 (44.0) | 8 (40.0) | 1.00b |
| Race, n (%) | .31c | |||
| White | 36 (80.0) | 18 (72.0) | 18 (90.0) | |
| African American | 6 (13.3) | 5 (20.0) | 1 (5.0) | |
| Other, non-Hispanic | 2 (4.4) | 1 (4.0) | 1 (5.0) | |
| Unknown | 1 (2.2) | 1 (4.0) | 0 (0.0) | |
| Location | ||||
| Colon/rectum | 13 (28.8) | 13 (52.0) | 0 (0.0) | |
| Appendix | 26 (57.7) | 6 (24.0) | 20 (100.0) | |
| Unknown | 6 (13.3) | 6 (24.0) | 0 (0.0) | |
P value calculated using Student t test.
P value calculated using Fisher exact test.
P value calculated using chi-square test.
All of the adenocarcinomas were labeled by FITC-PNA, versus 50% of the low-grade neoplasms (P < .01) (Table 2, Supplementary Figure 1, Supplementary Figure 2, Supplementary Figure 3, Supplementary Figure 4). The other lectin-binding results are in Table 2.
Table 2.
Summary of Lectin Binding Characteristics Using 5 FITC-Lectins
| All | Type of neoplasm |
P valuea | ||
|---|---|---|---|---|
| Adenocarcinoma | Low-grade | |||
| PNA | <.01 | |||
| Positive | 35 (77.8) | 25 (100.0) | 10 (50.0) | |
| Negative | 10 (22.2) | 0 (0.0) | 10 (50.0) | |
| RCA-1 | .03 | |||
| Positive | 22 (48.9) | 8 (32.0) | 14 (70.0) | |
| Negative | 22 (48.9) | 16 (64.0) | 6 (30.0) | |
| Unknown | 1 (2.2) | 1 (4.0) | 0 (0.0) | |
| SBA | 1.00 | |||
| Positive | 18 (40.0) | 10 (40.0) | 8 (40.0) | |
| Negative | 27 (60.0) | 15 (60.0) | 12 (60.0) | |
| UEA-1 | .18 | |||
| Positive | 33 (73.3) | 16 (64.0) | 17 (85.0) | |
| Negative | 12 (26.7) | 9 (36.0) | 3 (15.0) | |
| DBA | .01 | |||
| Positive | 29 (64.4) | 12 (48.0) | 17 (85.0) | |
| Negative | 16 (35.6) | 13 (52.0) | 3 (15.0) | |
DBA, Dolichos biflorus agglutinin; FITC, fluorescein isothiocyanate; PNA, peanut agglutinin; RCA-1, Ricinus communis agglutinin-1; SBA, soybean agglutinin; UEA-1, Ulex europeus agglutinin-1.
P values calculated using Fisher exact test.
Supplementary Figure 1.
Strongly positive peanut agglutinin staining within mucin deposition of an invasive mucinous adenocarcinoma (×10 objective, Zeiss epifluorescent microscope). The patient was later discharged to hospice.
Supplementary Figure 2.
Negative peanut agglutinin staining within mucin deposition of a low-grade neoplasm (×10 objective Zeiss epifluorescent microscope). The 4′,6-diamidino-2-phenylindole counterstain highlights intact cell nuclei within the mucin.
Supplementary Figure 3.
Positive Ricinus communis agglutinin-1 staining within mucin deposition of a low-grade neoplasm (×10 objective, Zeiss epifluorescent microscope). The 4′,6-diamidino-2-phenylindole counterstain highlights intact cell nuclei within the mucin and background benign colonic epithelium.
Supplementary Figure 4.
Positive Dolichos biflorus agglutinin staining within mucin deposition of a low-grade neoplasm (×20 objective, Zeiss epifluorescent microscope). The 4′,6-diamidino-2-phenylindole counterstain highlights intact cell nuclei within the mucin and background benign colonic epithelium.
Mucus is the term for the viscoelastic substance coating the intestinal epithelium, and its principal nonaqueous component is mucin. Mucin apoproteins have a linear polypeptide chain in its central axis.2 The oligosaccharide side chains of mucins often terminate with sialic acids or sulfate groups, making them negatively charged and hydrophilic. It has been demonstrated that virtually all normal human colonic mucins have the same terminal sugars on their oligosaccharide side chains.1 Goblet cells in the upper portion of the colonic crypt synthesize mucin with a terminal sugar that is bound by the lectins DBA and soybean agglutinin (α-N-acetylgalactosamine). There is a gradient of labeling by DBA and soybean agglutinin as goblet cells differentiate and migrate up the colonic crypt. The goblet cell mucins at the bottom of the crypt terminate with a sugar recognized by Ricinus communis agglutinin-1 (β-galactose). The disappearance of the terminal β-linked Gal residues at the base of the crypt is perfectly matched by the appearance of terminal α-linked GalNAc residues at the top. However, CRCs lose the organized crypt structure and there is a change in the pattern of lectin binding. PNA binds to the mucins made by almost all CRCs but essentially never binds to the mucins made in normal colons.3 PNA binds to the Thomsen-Friedenreich antigen (T-Ag), a cancer-associated sequence that consists of a short disaccharide pair (Galβ1-3->GalNAc).4 In some instances, the Gal, GalNAc of T-Ag, or both may be extended by terminal α-linked sialic acid residues; PNA also binds sialylated T antigens.4 Importantly, the presence of the PNA-binding glycoproteins can also be seen within large benign adenomatous polyps in foci of high-grade dysplasia.3
In this study, all of the mucinous adenocarcinomas were PNA-positive, indicating a cancer-associated T-Ag, whereas only half of low-grade neoplasms were PNA-positive. The other lectins were not as helpful diagnostically but shed light on the likely oligosaccharide structures in PMP (Table 2). One possibility is that terminal sialic acid residues are permissive of PNA-binding but inhibit Ricinus communis agglutinin-1 binding. DBA binds the terminal α-GalNAc of blood group A and labeling was present in about half of the adenocarcinomas but most of the low-grade neoplasms. It is likely that DBA recognizes different GalNAc residues in the cancers than in the normal colon, because in addition to recognizing the blood group A structure, DBA also binds to solitary α-linked GalNAc, such as that in the cancer-associated structure, Tn.5 The presence of Tn and sialylated-Tn in mucins is associated with metastasis and a poorer prognosis.6
It has been previously shown that CRC-associated mucins are less densely glycosylated, have shorter oligosaccharide side chains than in the normal colon,7 and have other biochemical differences in molecular weight and charge.8 Future directions include purification and biochemical characterization of these cancer-associated mucins, and exploration of the molecular and enzymatic basis of this phenomenon.
Acknowledgments
L. August Clark conducted the fluorescence microscopy, assembled and interpreted data, and wrote the manuscript. Alexia Ghazi performed immunofluorescence interpretation and photographs, patient data collection, and edited the manuscript. Kristin Gaffney initiated the experiments, collected the samples, and conducted initial fluorescence microscopy. Rodrigo Soto performed initial immunofluorescence interpretation and photographs. Atin Agarwal edited the manuscript. Susanne Carmack read all of the PMP tissues blinded to the lectin-binding data. C. Richard Boland conceived of the experiments, assisted with fluorescence microscopy, and edited the manuscript.
Footnotes
Conflicts of interest The authors disclose no conflicts.
Supplementary Material
References
- 1.Boland C.R., Montgomery C., Kim Y. Alterations in human colonic mucin occurring with cellular differentiation and malignant transformation. Proc Natl Acad Sci USA. 1982;79:2051–2055. doi: 10.1073/pnas.79.6.2051. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Byrd J., Bresalier R. Mucins and mucin binding proteins in colorectal cancer. Cancer Metastasis Rev. 2004;23:77–99. doi: 10.1023/a:1025815113599. [DOI] [PubMed] [Google Scholar]
- 3.Boland C.R., Montgomery C., Kim Y. A cancer-associated mucin alteration in benign colonic polyps. Gastroenterology. 1982;82:664–672. [PubMed] [Google Scholar]
- 4.Bian C.F., Zhang Y., Sun H., Li D.F., Wang D.C. Structural basis for distinct binding properties of the human galectins to Thomsen-Friedenreich antigen. PLoS One. 2011;6:e25007. doi: 10.1371/journal.pone.0025007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Barrow H., Tam B., Duckworth C.A., Rhodes J.M., Yu L.G. Suppression of core 1 Gal-transferase is associated with reduction of TF and reciprocal increase of Tn, sialyl-Tn and core 3 glycans in human colon cancer cells. PLoS One. 2013;8:e59792. doi: 10.1371/journal.pone.0059792. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Munkley J. The role of sialyl-Tn in cancer. Int J Mol Sci. 2016;17:275. doi: 10.3390/ijms17030275. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Boland C.R., Deshmukh G. The carbohydrate composition of mucin in colonic cancer. Gastroenterology. 1990;98:1170–1177. doi: 10.1016/0016-5085(90)90330-4. [DOI] [PubMed] [Google Scholar]
- 8.Shimamoto C., Deshmukh G.D., Rigot W.L., Boland C.R. Analysis of cancer-associated colonic mucin by ion-exchange chromatography: evidence for a mucin species of lower molecular charge and weight in cancer. Biochem Biophys Acta. 1989;991:284–295. doi: 10.1016/0304-4165(89)90118-9. [DOI] [PubMed] [Google Scholar]