Elevated MUC5AC expression is associated with mismatch repair deficiency and proximal tumor location but not with cancer progression in colon cancer

Mucin 5AC (MUC5AC) is a secreted gel-forming mucin expressed by several epithelia. In the colon, MUC5AC is expressed in scattered normal epithelial cells but can be abundant in colorectal cancers. To clarify the relationship of MUC5AC expression with parameters of tumor aggressiveness and mismatch repair deficiency (dMMR) in colorectal cancer, a tissue microarray containing 1812 colorectal cancers was analyzed by immunohistochemistry. MUC5AC expression was found in 261 (15.7%) of 1,667 analyzable colorectal cancers. MUC5AC expression strongly depended on the tumor location and gradually decreased from proximal (27.4% of cecum cancers) to distal (10.6% of rectal cancers; p < 0.0001). MUC5AC expression was also strongly linked to dMMR. dMMR was found in 21.3% of 169 cancers with MUC5AC positivity but in only 4.6% of 1051 cancers without detectable MUC5AC expression (p < 0.0001). A multivariate analysis showed that dMMR status and tumor localization predicted MUC5AC expression independently (p < 0.0001 each). MUC5AC expression was unrelated to pT and pN status. This also applied to the subgroups of 1136 proficient MMR (pMMR) and of 84 dMMR cancers. The results of our study show a strong association of MUC5AC expression with proximal and dMMR colorectal cancers. However, MUC5AC expression is unrelated to colon cancer aggressiveness. Supplementary Information The online version contains supplementary material available at 10.1007/s00795-020-00274-2.


Introduction
Colorectal cancer was the third most common cancer worldwide in 2018 and the second most common cause for cancer related death [1]. Standard treatment of colorectal cancer consists of surgical removal. In high-risk cancers, adjuvant chemotherapy is also given in order to destroy micrometastasis and to reduce the risk of local recurrence. Possible chemotherapies include conventional cytotoxic chemotherapy and several antiangiogenic substances. In case of BRAF, KRAS, and NRAS wild-type cancers, anti-EGFR therapy antibodies can also be applied (summarized in [2]). Immune checkpoint inhibitors can be administered in cancers harboring microsatellite instability (MSI) or mismatch repair deficiency (dMMR) (summarized in [3]). Established prognostic factors of colorectal carcinomas include pT, pN, M status and histologic tumor features [4,5]. They are statistically powerful but cannot reliably predict disease course in individual patients.
None of the earlier studies on the putative clinical role of MUC5AC expression in colon cancer have considered the association between MUC5AC expression and clinicpathological parameters in the subgroups of mismatch repair proficient (pMMR) and dMMR cancers in their analyses. We thus analyzed the relationship between MUC5AC expression and features of tumor aggressiveness (pT and pN) in all cancers and in the subgroup of pMMR and dMMR cancers in a cohort of 1812 colorectal cancers by immunohistochemistry (IHC) in a tissue microarray (TMA) format.

Tissue microarray (TMA)
Our colon cancer TMA consisted of 1,812 colon cancers diagnosed at the Institutes of Pathology of the University Medical Center Hamburg-Eppendorf (Hamburg, Germany) and the Department of Pathology of the Academic Hospital Fuerth (Fuerth, Germany) between 2009 and 2019. TMA construction was done as previously described [40]. Clinical, pathological and molecular parameters were obtained from patient records (Table 1). The use of archived remnants of diagnostic tissues for manufacturing of tissue microarrays and their analysis for research purpose as well as patient data analysis has been approved by local laws (HmbKHG, §12) and by the local ethics committee (Ethics Commission Hamburg, WF-049/09). All work has been carried out in compliance with the Helsinki Declaration.

Immunohistochemistry (IHC)
Freshly prepared TMA sections were immunostained on one day in one experiment. Slides were deparaffinized and exposed to heat-induced antigen retrieval for 5 min in an autoclave at 121 °C in pH 7.8 Dako Target Retrieval Solution buffer (Dako, Glostrup, Denmark). Primary antibody specific against MUC5AC protein (mouse monoclonal, MSVA-109, MS Validated Antibodies, Hamburg, Germany) was applied at 37 °C for 60 min at a dilution of 1:200. Bound antibody was then visualized using the EnVision Kit (Dako, Glostrup, Denmark) according to the manufacturer's instructions. For immunostaining of the MMR proteins, primary antibodies (ready to use) specific for MLH1 (clone ES05, mouse), PMS2 (clone EP51, rabbit), MSH2 (clone FE11, mouse), and MSH6 (clone EP49, rabbit) from Dako (Glostrup, Denmark) were applied for 20 min (MLH1, MSH2, MSH) or 30 min (PMS2) in an automated immunostainer (Dako/Agilent Autostainer Link 48 (Santa Clara, USA). MUC5AC staining was seen in the membrane and cytoplasm of the colon cancer cells and immunostaining was interpreted as follows: Negative: no staining at all tumor cells, weak staining: staining intensity of 1 + in ≤ 70% of the tumor cells or staining intensity of 2 + in ≤ 30% of the tumor cells, moderate staining: staining intensity of 1 + in > 70% of the tumor cells, staining intensity of 2 + in > 30% but in ≤ 70% of the tumor cells

Technical issues
MUC5AC expression was informative in 1667 (92%) of the 1,812 arrayed cancers in our IHC analysis. Reasons for non-informative cases (n = 145; 8%) included lack of tissue samples or the absence of unequivocal cancer cells in the TMA spot.

MUC5AC expression and tumor phenotype
In normal colorectal epithelial cells, MUC5AC expression was only seen in few scattered epithelial cells (Fig. 1a). In colorectal cancer, positive MUC5AC staining was seen in 261 (15.7%) of 1667 analyzable tumor spots. The staining patterns included variable numbers of interspersed positive cells, patchy focal staining, and intense diffuse positivity ( Fig. 1b- Table 2). This association was independent of the tumor localization, as the frequency of dMMR cancers is higher in proximal and distal colorectal tumors with MUC5AC positive in comparison with MUC5AC-negative cancers. Lack of significance in some subgroups may be due to low case numbers (Table 3). In addition, our multivariate analyses showed that firstly MUC5AC expression and tumor localization independently predicted dMMR status and secondly dMMR status and tumor localization independently predicted MUC5AC expression status (Table 4). In the subgroups of dMMR and pMMR cancers, MUC5AC expression was unrelated to pT and pN status (Table 2).

Discussion
MUC5AC expression was found in 16% of colorectal cancers in our study. This is in the lower range of data from previous studies. A total of 23 studies have earlier analyzed MUC5AC expression in colorectal cancer (  [30], and Biemer-Hüttmann et al. in 55% of 40 proximal and in 22% of 23 rectal cancers [18]. A gradual change of biomarker expression from proximal to distal colon could, for example, be explained by continuing changes in the density and the composition of the stool, and the exposure time to different carcinogenic factors during the colon passage or be related to the embryonal development of the colon (summarized in [41]). Other examples of biomarkers that were described to vary dependent on the localization within the colon, for example, include AMACR [42], p53 [30] as well as amplification of EGFR and HER2 [43].
At least 9 studies have analyzed MUC5AC expression and dMMR/MSI in colon cancer, and all of them have described significant associations (Supplementary Table 3 [18]. Given the wellknown associations of MSI with right-sided colon cancer location (reviewed in [44]), it was expected that MUC5AC (as any biomarker) would be linked to both or none. However, our multivariate analysis revealed that tumor location and dMMR independently enhanced the likelihood for detectable MUC5AC expression in colorectal cancer. The underlying molecular mechanism is not known. However, one study hypothesizes an increased likelihood of MUC5AC promotor hypomethylation in MSI cancers. MUC5AC promotor hypomethylation results in MUC5AC upregulation and was shown to be associated with a "mutator" phenotype in-especially MSI-colon cancers [45].
In a thorough experimental study, Pothuraju et al. have recently shown that differential MUC5AC expression drives tumorigenesis and promotes aggressiveness of colorectal cancer cells in vitro and in vivo. The authors examined the impact of reduced (shRNA-mediated) or absent (CRISPR/Cas9-mediated) MUC5AC expression on cell proliferation, anchorage independent cell growth, cell migration, and cell invasion in two endogenous MUC5AC expressing colorectal cancer cell lines. In addition, they prepared a MUC5AC knockout xenograft model to investigate the impact of MUC5AC on tumorigenesis in vivo.

Fig. 2 MUC5AC expression and colorectal cancer localization
Overall, the authors showed that MUC5AC expression enhanced cell growth, invasion and migration and decreased apoptosis in vitro and increased tumorigenesis in vivo [38]. However, the absence of associations between MUC5AC expression and pT as well as pN argues against a clinically relevant role of MUC5AC expression for driving aggressiveness of human colorectal cancer cells. A lack of clinical relevant prognostic impact of MUC5AC expression is in line with the conflicting results of 11 earlier studies analyzing the clinical relevance of MUC5AC expression in colorectal cancer (summarized in Table 5). Of these, seven studies with 35-381 patients described an association of high MUC5AC expression with favorable phenotype and/or prognosis [12,17,20,21,23,31,35], three studies with 33-250 patients reported a link between high MUC5AC expression and poor phenotype and/or prognosis [25,34,35], and five studies involving 35-206 patients could not find any relationship between MUC5AC expression and clinic-pathological features [11,12,18,21,31]. The fact that MUC5AC expression was also unrelated to aggressive cancer phenotype in our 1051 pMMR cancers demonstrates that adverse prognostic effects of MUC5AC are not obscured by the favorable prognostic influence of dMMR.
It is of note that MUC5AC may also represent a suitable drug target. Ensituximab (Neo-102), a novel chimeric monoclonal antibody, binds to an aberrantly glycosylated cancer-associated MUC5AC variant and is able to activate the immune system to exert a cytotoxic T-lymphocyte response [47]. In a phase I study of pancreatic cancer patients preselected for MUC5AC expression, a favorable toxicity profile was found for Ensituximab [47]. In a recent phase II study, Ensituximab resulted in stable disease in 21% of 56 patients with heavily pretreated refractory colorectal cancers and was well tolerated [48]. Given the high numbers of inflammatory cells occurring in MSI/ dMMR-positive colorectal cancers, one might speculate that Ensituximab treatment might be potentially promising in these carcinomas. In summary, the results of our study show that elevated MUC5AC expression is independently linked to proximal location and dMMR in colorectal cancers. However, both in dMMR and in pMMR cancers, MUC5AC expression is unrelated to aggressive cancer phenotype.

Competing interest
The Institute of Pathology of the UKE receives royalties on the sales of MUC5AC clone MSVA-109 from MS Validated Antibodies GmbH (owned by a family member of GS).

Ethical approval
The usage of archived diagnostic left-over tissues for manufacturing of tissue microarrays, their analysis for research purposes and patient data analysis has been approved by local laws (HmbKHG, §12,1) and by the local ethics committee (Ethics commission Hamburg, WF-049/09). All work has been carried out in compliance with the Helsinki Declaration.