Development and Characterization of 3-Dimensional Cell Culture Models of Adrenocortical Carcinoma Authors: Sarah Feely, Nathan Mullen, Padraig T Donlon, Eileen Reidy, Ritihaas Surya Challapalli, Mariam Hassany, Anna Sorushanova, Eduardo Ribes Martinez, Peter Owens, Anne Marie Quinn, Abhay Pandit, Brendan Harhen, David P Finn, Constanze Hantel, Martin O’Halloran, Punit Prakash, Michael C Dennedy Published: 2024 ABSTRACT AbstractAdrenocortical carcinoma (ACC) is a rare malignancy of the adrenal cortex that is associated with a poor prognosis. Developing effective treatment options for ACC is challenging owing to the current lack of representative preclinical models. This study addressed this limitation by developing and characterizing 3-dimensional (3D) cell cultures incorporating the ACC cell lines, MUC-1, HAC15, and H295R in a type I collagen matrix. ACC tissue samples were analyzed by immunohistochemistry to determine the presence of type I collagen in the tumor microenvironment. Cell viability and proliferation were assessed using flow cytometry and confocal microscopy. mRNA expression of steroidogenic enzymes and steroid secretion was analyzed by comparing the 3D and monolayer cell culture models. All cells were successfully cultured in a type I collagen matrix, which is highly expressed in the ACC tumor microenvironment and showed optimal viability until day 7. All 3 models showed increased metabolic and proliferative activity over time. Three-dimensional cell cultures were steroidogenic and demonstrated increased resistance to the gold standard chemotherapy, mitotane, compared with monolayer. The use of these models may lead to an improved understanding of disease pathology and provide a better representative platform for testing and screening of potential therapies.Keywords: 3D cell culture models, adrenocortical carcinoma, preclinical models, drug resistance Adrenocortical carcinoma (ACC) is a rare and aggressive malignancy, with an incidence of 1 to 2 per million people annually and a poor prognosis, with a median overall survival of approximately 24 months (1-3). Most cases are diagnosed at an advanced stage with local spread or metastases (4, 5). The only curative option is R0 surgical resection of limited disease, ENS@T Stage 1 to 2 (1, 6). Recurrent disease is typically managed with adjuvant mitotane or systemic chemotherapy (cisplatin/etoposide/doxorubicin or streptozocin), although both regimens have low recurrence-free and overall survival rates (7, 8). ACC presents several research and clinical challenges. Chemotherapy response rates are poor and obtaining clinical samples or primary cells for research is difficult due to its rarity (7, 9-15). Moreover, primary cells do not adhere or proliferate easily in cell culture (9-15). While several steroidogenic ACC cell lines, such as MUC-1, H295R, and HAC15, have become available, only 2 are commercially accessible (H295R and HAC15) (16-22). Our laboratory works with these 3 cell lines, including MUC-1, which is notable for its resistance to mitotane (18). Reliable preclinical ACC models remain limited, and currently rely mostly on monolayer cell culture or xenografts (14, 17, 18, 23-32). There is therefore a pressing need for improved models, preferably animal-sparing, to advance drug screening and understand mechanisms of cell survival and death in ACC. ACC tumors are large, typically over 6 cm, often necrotic, highly proliferative (Ki67-high), and steroidogenic (5, 33-35). Hormonal excess, particularly cortisol, leads to Cushing syndrome, worsening prognosis due to complications like hypertension and immunosuppression (36). Androgen excess is also common, while mineralocorticoid excess is rare (37). Radiological diagnosis of ACC is aided by biochemical testing of steroid excess including urine steroid metabolomics, which detect adrenocortical steroid precursors (38-41). Histologically, ACC is evaluated using the Weiss Score and Ki67 score, both important prognostic markers, where higher scores indicate higher recurrence and worse outcomes overall (1, 33, 42). Preclinical models of ACC should reflect the disease's heterogeneity, including steroidogenesis, varied proliferation rates, central necrosis, and chemotherapy resistance. Current research still relies heavily on monolayer cell cultures (H295R and HAC15), which fail to fully mimic the tumor microenvironment and are limited in studying key interactions between tumor cells, stroma, and the extracellular matrix (ECM) (43, 44). Monolayer cell cultures also lack areas of confluent necrosis and do not effectively model invasion or migration (45, 46). In contrast, 3-dimensional (3D) cell culture models better replicate the tumor microenvironment, exhibiting more representative growth patterns, necrosis, and hypoxia, and allow for modeling of increasing complexity which can mimic interactions between cancer cells and immune, stromal, and stem cells (46-50). Complex microfluidic systems can also be used to further model vascular invasion and the endothelial barrier and to improve the translational relevance of these models for drug screening (51). While recent advances have led to the development of 3D spheroid and coculture models for ACC, a significant knowledge gap remains. Few models incorporate ECM components or complex multicellular interactions (28, 52-63). Previous studies have used only Matrigel and hyaluronic acid/gelatin to model the ECM and there has been no previously published work which has modeled collagen as a more biologically relevant ECM (55, 57, 59). In the current study we address this knowledge gap by using type I collagen, abundant in ACC, to offer a more translationally relevant system for evaluating tumor response to therapy with the aim of providing an animal-sparing approach to drug testing. We have specifically developed a 3D cell culture model for ACC which uses 3 human ACC cell lines embedded in a type I collagen matrix which is cross-linked to represent tumor stiffness. We describe a model which offers the potential for the addition of greater complexity to incorporate stromal and immune elements to better represent the tumor microenvironment as an animal sparing preclinical drug screening model of ACC. Materials and Methods Clinical Sample Collection Paraffin-embedded tumor tissue and clinical details were collected from patients undergoing adrenalectomy at University Hospital Galway between 2014 and 2024, including samples from ACCs, adrenal adenomas, and pheochromocytomas. All samples were obtained with informed consent, and ethics approval was provided by the University Hospital Galway Research Ethics Committee (ref. C.A. 3194). Immunohistochemistry Collagen expression in normal and tumor tissues was evaluated via immunohistochemistry. Nonspecific binding was blocked using horse serum (Vector Laboratories, catalog # S-2012-50) for 1 hour and sections were stained with a type I collagen antibody (Thermofisher, catalog # PA1-26204, RRID:AB_2260734). Slides were incubated with a horseradish peroxidase-conjugated secondary antibody (Vector Laboratories, catalog # MP-7500, RRID:AB_2336534) and visualized with DAB (Vector Laboratories, catalog # SK-4105, RRID:AB_2336520) before hematoxylin counterstaining. A histopathologist, blinded to diagnosis, reviewed all slides. Cell Culture Monolayer H295R (primary ACC tumor) (NCI-H295R, RRID:CVCL_0458), HAC15 (H295R subclone) (CRL-3301, RRID:CVCL_S898), and MUC-1 (metastatic ACC) cell lines were maintained in different media specific to each line (16, 18-22). Cells were cultured in a humidified atmosphere with 5% CO2 at 37 °C and used between specified passages. All cell lines were validated using short tandem repeat (STR) profiling. Three-dimensional cell culture Three-dimensional cell culture models were generated by embedding cells in a bovine type I collagen matrix(Collagen Solutions, catalog # FS22005), mixed with 4arm PEG succinimidyl glutarate (pentaerythritol) (Jenkem USA, catalog # A7031) and cells, forming a gel. Each cell line was seeded in the collagen matrix and incubated at 37 °C with 5% CO2. The composition of the 3D culture matrix was based on previously optimized concentrations (unpublished work). Cellular viability Cell viability in 3D cell cultures was assessed using Calcein-AM (Merck, catalog # 17783) (8 mM) and propidium iodide (Miltenyi Biotec, catalog # 130-093-233) (5 µg/mL) and visualized with confocal microscopy. Flow cytometry using Sytox blue (Invitrogen catalog #S 34857) was employed for quantitative analysis of cell death, with monolayer and 3D cell cultures dissociated and stained at days 3, 7, 14, and 21. Flow cytometry data was analyzed using De Novo software FCS express V 7.14.0020 or FlowJo v10.8 software (BD Life Sciences). The gating strategy is illustrated elsewhere (Fig. S1 (64)). Proliferation and metabolic activity Ki67 antibody (Biolegend, catalog # 151210, RRID:AB_2716008) staining was used to evaluate proliferation in both monolayer and 3D cell cultures, followed by flow cytometry. Metabolic activity was assessed using alamarBlue (ThermoFisher, catalog # DAL1025) fluorescence assays, comparing the activity of monolayer and 3D cell cultures spectrophotometrically excited at a wavelength of 590/40 nm using the Hidex Sense Microplate Reader. Steroid quantification Steroid secretion (aldosterone, cortisol, and androstenedione) was stimulated with 10 nM angiotensin II (Ang II) (Merck, catalog # A9525-1MG) and 10 μM forskolin (FSK) (Merck, catalog # F6886, 10MG) for 24 hours in monolayer and 3D cell cultures, with media supernatants collected and stored for analysis as previously described (16, 65). Steroid levels were quantified via high-performance liquid chromatography (HPLC) coupled to tandem mass spectrometry, with internal standards and standard curves used for concentration calculations. Gene expression Semiquantitative reverse transcription polymerase chain reaction was used to measure steroidogenic enzyme expression (CYP11B1, CYP11B2, and StAR) as previously described (65). RNA was isolated from monolayer and 3D cell cultures, reverse-transcribed, and analyzed using real-time polymerase chain reaction with SYBR green detection. Analysis was performed using the ΔΔCT method on the QBase + analytical software, Version 3.3 (Qiagen), against a geometric mean of 2 reference genes (RPLPO and PMM1) in monolayer. Data are represented as the fold change in mRNA relative to the monolayer cell culture control at day 7. Drug treatments Cells in monolayer and 3D cell cultures were treated with various concentrations of mitotane (Merck, catalog #SML1885) for 24 or 48 hours, respectively. A DMSO (Merck, catalog # D2650) vehicle control was included, and treatment was administered after 5 days of 3D cell culture. Statistical Analysis Data are presented as mean ± SEM. Paired sample analyses were conducted using 1-way analysis of variance (ANOVA), with multiple-group comparisons analyzed using 2-way ANOVA and post hoc tests. Statistical significance was set at P < .05 unless stated otherwise, indicated on graphs as *P < .05, **P < .01, and ***P < .001. Detailed methodology is available elsewhere (Supplementary Materials section 1 (64)). Results ACC Cells Express Type I Collagen in Their Extracellular Matrix Type I collagen is 1 of the most highly expressed proteins in the tumor microenvironment in various cancer types and represents the most suitable matrix for generating 3D cancer models (66). To verify its suitability for 3D cell culture of ACC, we measured the expression of type I collagen in human ACC and benign adrenal adenomas using immunohistochemistry. Type I collagen was strongly expressed in the tumor microenvironment of ACC (Fig. 1A). There was no difference in staining intensity between the ACC tissues and control skin tissue samples (Fig. 1A). Type I collagen was also equally expressed in the tumor stroma of non-aldosterone–producing adenomas, phaeochromocytomas, and aldosterone producing adenomas, with no significant difference between each tissue type (Fig. 1B). However, expression of type I collagen was not observed in normal adrenal tissue. ACC cells express type I collagen in their extracellular matrix. (A) Graphical representation of mean staining intensity in adrenocortical carcinoma (ACC) tissues (n = 4) vs skin (n = 3). (B) Graphical representation of the mean staining intensity in ACC tissues vs non-aldosterone–producing adenoma (NAPACA) (n = 2), phaeochromocytoma (PC/PGL) (n = 3), and aldosterone-producing adenoma (APA) (n = 3) tissues. (C) Immunohistochemistry images of skin, ACC, NAPACA, phaeochromocytoma, and APA. Data are represented as the mean ± SEM. Statistical comparisons were performed using (1) unpaired t test (skin vs ACC) and (2) 1-way ANOVA followed by post hoc analysis when comparing the expression between adrenal tissues. ACC Cells in 3D Cell Culture Maintain Viability Over 21 Days in Culture Next, we evaluated the viability of each ACC cell line (MUC-1, H295R, and HAC15) in 3D cell culture on days 7 and 14 using confocal imaging and on days 3, 7, 14, and 21 using flow cytometry. Flow cytometry analysis demonstrated that all 3 cell lines were viable in 3D cell culture, with the highest viability observed during the first 7 days. The viability of H295R and MUC-1 cells decreased significantly at days 14 and 21 (Fig. 2A and 2C), while HAC15 cells demonstrated lower baseline viability than the other cell lines but remained stable over 21 days (Fig. 2B). Confocal microscopy confirmed cell viability in all 3 cell lines (Fig. 2D), though an increase in the number of dead cells was observed over time. H295R cells showed areas of confluent necrosis on days 7 and 14. Interestingly, HAC15 cells demonstrated a necrotic core in 3D cell culture. However, this was not observed when the cells were analyzed by flow cytometry, indicating that the necrotic core is lost upon extraction from the collagen matrix (Fig. 2D). ACC cells in 3D cell culture maintain viability over 21 days in culture: (A-C) Cell viability in H295R, HAC15, and MUC-1 cells in 3D cell culture vs monolayer cell culture measured using flow cytometry analysis of Sytox Blue staining. Representative histograms and gating strategies are presented elsewhere (Fig. S1 (64)). (D) Representative confocal microscopy images using calcein (green) and propidium iodide (red/no stain) demonstrated high levels of cell viability in all 3 cell lines at day 7. HAC15 cells showed a visible necrotic core on day 14. Scale bar: 500 mm. Data are represented as the mean ±SEM. Statistical comparisons were performed using 2-way ANOVA followed by post hoc analysis compared to monolayer. Statistical significance is denoted as *P < .05, **P < .01, ****P < .0001. When comparing the viability of cells in 3D cell cultures vs monolayer cell cultures, H295R and HAC15 cells showed lower viability in 3D cell culture at all time points (Fig. 2E and 2F). MUC-1 cells also showed lower viability in 3D cell culture, but only at days 14 and 21 (Fig. 2E). H295R and HAC15 ACC Cells are Steroidogenic When Cultured in 3D Cell Culture Models Steroidogenesis occurs in most ACC patients in vivo (35, 38). In monolayer cell culture, HAC15 and H295R ACC cell lines are well described as steroidogenic at baseline and under stimulated conditions (65). Steroidogenesis in MUC-1 cells is less well characterized with conflicting results in the literature (16, 18). We investigated whether each cell line demonstrated a steroidogenic milieu in monolayer and in 3D cell culture by measuring the secretion of cortisol, androstenedione, and aldosterone, as well as mRNA expression of steroidogenic enzymes/mediators. Under FSK stimulated conditions, on 1 hand, H295R and HAC15 cells secreted significantly more cortisol in monolayer than 3D cell culture for most timepoints measured (Fig. 3A and 3B). MUC-1, on the other hand, did not demonstrate steroidogenesis for cortisol or aldosterone in either monolayer or 3D cell culture under stimulated or unstimulated conditions. Androstenedione was present in all 3 cell lines under stimulated conditions and was either similar (H295R at all timepoints; HAC15 D14, MUC-1 D14) or higher (HAC15 D7, MUC-1 D7) in 3D cell culture than monolayer cell culture. (Figure 3F-3H). Under Ang II stimulation, aldosterone secretion was significantly higher in 3D cell culture for H295R (D14) and HAC15 (all timepoints) cells than in monolayer (Fig. 3D and 3E). ACC cells are steroidogenic when cultured in 3D cell culture models: cortisol, aldosterone and androstenedione secretion following stimulation with FSK (10 mM) (cortisol and androstenedione secretion), or with ANGII (10 nM) (aldosterone secretion) measured using HPLC tandem mass spectrometry cortisol secretion for H295R (A) and HAC15 (B). Aldosterone secretion for H295R (C) and HAC15 (D). Androstenedione secretion for H295R (E), HAC15 (F), and MUC-1 (G). Data are represented as the mean ±SEM. Statistical comparisons were performed using 2-way ANOVA followed by post hoc analysis compared with monolayer. Statistical significance is denoted as *P < .05, **P < .01, ***P < .001, ****P < .0001. Expression of key steroidogenic enzymes, including CYP11B1 and CYP11B2, was measured under stimulated conditions for all cell lines in monolayer and 3D cell culture following 7 and 14 days in cell culture. No significant difference was demonstrated for the expression of these enzymes or StAR protein between monolayer and 3D cell culture or across the measured timepoints (Fig. S2A-H (64)). In MUC-1 cells, there was no expression of either enzyme in line with the absence of steroidogenesis for cortisol and aldosterone. StAR expression was seen however, with a higher expression in monolayer cell culture compared to 3D cell culture at day 14 only (Fig. S2I (64)). ACC Cells Proliferate and Have High Metabolic Activity in 3D Cell Culture Proliferation and metabolic activity of ACC cells were evaluated in monolayer and 3D cell cultures. Metabolic activity increased in all 3 cell lines over time, with the highest fluorescence readings observed at day 14 (Fig. 4A-4C). When comparing monolayer and 3D cell cultures, all 3 cell lines demonstrated lower metabolic rates in 3D cell culture, except for MUC-1 at day 21 (Fig. 4D-4F). ACC cells proliferate and have high metabolic activity in 3D cell culture. (A-C) Metabolic activity in H295R, HAC15, and MUC-1 cells in 3D cell culture measured using alamarBlue staining. (D-F) Comparison between metabolic activity in monolayer vs 3D cell culture for each cell line measured using alamarBlue staining. Ki67 positive (+) and Ki67 (−) expression in (G and J) H295R, (H and K) HAC15, and (I and L) MUC-1 3D and monolayer cell culture models, respectively. Data are represented as mean±SEM. Statistical comparisons were performed using 2-way ANOVA followed by post hoc analysis. Statistical significance is denoted as *P < .05, **P < .01, ***P < .001, ****P < .0001. Data are representative of 3 independent experiments. Cellular proliferation was measured using Ki67 expression, with all 3 cell lines showing a broadly similar proliferation pattern between monolayer and 3D cell cultures. H295R cells in 3D cell culture showed a higher proportion of actively proliferating cells across all time points (Fig. 4G). In HAC15 cells, a higher proportion of proliferating cells was observed at day 14 and 21 in 3D cell culture (Fig. 4H), while MUC-1 cells showed greater proliferation in 3D cell culture at days 7 and 14 (Fig. 4I). In contrast, monolayer cell culture showed a decrease in proliferation over time in MUC-1 and H295R cell lines (Fig. 4J and 4L). This decrease was only seen at day 7 in HAC15 cells (Fig. 4K). ACC Cells are Resistant to Mitotane When Treated in 3D Cell Culture Compared With Monolayer Cell Culture Previous studies have demonstrated that cancer cells grown in 3D cell cultures display greater chemotherapy resistance than monolayer cell cultures (67, 68). In this study, the response of each ACC cell line to mitotane was evaluated in 3D cell culture and compared with monolayer cell culture. In monolayer cell culture, H295R cells were more sensitive to mitotane than HAC15 or MUC-1 cells (Fig. 5). However, in 3D cell culture, lethal doses of mitotane displayed significantly lower cytotoxicity across all cell lines, requiring much higher doses to achieve cell death (Fig. 5A-5C). Cell death in 3D cell cultures was equally distributed throughout the collagen matrix (Fig. S3 (64)). ACC cells are resistant to mitotane when treated in 3D cell culture compared to monolayer cell culture: (A-C) % live cells of H295R, HAC15, and MUC-1 cells in 3D cell culture vs monolayer measured using flow cytometry analysis of Sytox blue staining. Data are represented as mean ± SEM. Statistical comparisons were performed using 2-way ANOVA followed by post hoc analysis compared with monolayer. Statistical significance is denoted as *P < .05, ****P < .0001. Data are representative of 3 independent experiments. Discussion In this study, we developed and characterized a novel 3D cell culture model of ACC using 3 steroidogenic ACC cell lines (H295R, HAC15, and MUC-1) embedded in a type I collagen ECM. We validated the use of type I collagen by confirming its expression in the ECM of human adrenocortical tumors, including ACC. The 3D cell culture model was analyzed for viability, steroidogenesis, proliferative activity, and chemotherapy response. While baseline cellular viability in 3D cell culture was lower than in monolayer cell culture, proliferation was comparable, indicating high cell turnover. H295R and HAC15 cell lines remained steroidogenic in 3D cell culture, as demonstrated by the expression of steroidogenic enzymes and steroid secretion. In our hands the metastatic ACC model, MUC-1 cells were steroidogenic for androstenedione only and did not express CYP11B1 or CYP11B2. Importantly, the cells in 3D cell culture showed reduced mitotane-induced cell death compared to monolayer cell cultures, reflecting a more representative pharmacological response of ACC in the tumor microenvironment. ACC is a rare cancer with limited models available for research (44). While several steroidogenic ACC cell lines exist, only H295R and HAC15 are available commercially (16-22). More recently MUC-1 cells have provided a better in vitro model for chemotherapy-resistant metastatic disease, and in this study, we demonstrate that HAC15 cells also show mitotane resistance (18). Primary cell culture remains challenging, and clinical studies are difficult to recruit due to the low incidence of ACC (7, 9-15). Finally, few animal models of ACC exist, and most are reliant upon the use of human cell line xenografts (44, 69). Overall, current models of drug screening, especially with monolayer cultures, have not translated well to improved patient outcomes, demonstrating the need to explore novel and improved methodologies for drug screening in this disease (11, 70-72). Three-dimensional cell culture addresses some limitations of monolayer cell culture by offering a more representative tumor microenvironment that includes cancer cells, stromal cells, and the ECM (43). Collagen, particularly type I collagen, is the most abundant ECM protein, associated with tumor growth and invasion in other cancer models (73). However, this ECM protein had not previously been used to model 3D cell cultures in ACC, which favored the use of Matrigel and hyaluronic acid/gelatin (55, 57, 59). After verifying the expression of type I collagen in ACC, we determined that the bovine collagen ECM would be an ideal matrix for 3D ACC cell culture model, which was then used for 3 available cell lines which differed in their chemotherapy response and steroidogenic milieu. In our study, all 3 cell lines continued to proliferate and maintained steroidogenesis in 3D cell culture over 21 days, although demonstrating lower viability than monolayer cultures. The presence of confluent necrosis, particularly in HAC15 cells, mirrored the in vivo tumor environment. The reduced baseline viability observed in 3D cell cultures likely results from cellular overcrowding, leading to competition for nutrients and oxygen, particularly in the core of the cultures (67). This is consistent with the hypoxic and nutrient gradients that lead to necrosis in tumors (46, 74). While cortisol production was lower in 3D cell culture, there was higher aldosterone and androgens production than monolayer cell culture. This is consistent with a prior study of ACC cell cultures in a spheroid model (61). However, the mechanism for this has not yet been studied. Therefore, our overall findings show that 3D ACC cell culture models demonstrate numerous disease characteristics, making them useful for drug screening and studies of cytotoxic chemotherapy. The cytotoxicity of mitotane was tested in all 3 cell lines, showing the expected resistance in MUC-1 cells (75) and, for the first time, mitotane resistance in HAC15 cells in monolayer. In 3D cell cultures, all 3 cell lines exhibited lower sensitivity to mitotane than monolayer, aligning with results seen in other 3D cancer models including breast and bowel cancer (67, 68, 76). The enhanced resistance in 3D cell cultures highlights the advantage of using 3D cell culture models for drug screening, as they better replicate in vivo responses to chemotherapy, which are typically less robust than the response seen in monolayer cell culture (77). The use of 3D cell culture models to study cancer is increasingly recognized as a standard (43, 68). However, technical challenges remain, such as accurately measuring cell viability and proliferation within 3D cell cultures (78). In our study, we utilized both real-time imaging and flow cytometry to validate our results, ensuring consistency between intact gels and extracted cells. Further refinement of assays to minimize cell extraction will enhance future research using 3D cell culture models. In conclusion, we characterized a 3D cell culture model of ACC using 3 steroidogenic cell lines in type I collagen, which retained key characteristics of ACC. This model demonstrated greater resistance to mitotane cytotoxicity than monolayer cultures, offering a more complex and representative model for investigating ACC. We propose that this model could serve as an animal-sparing platform for drug screening and for studying the mechanisms of chemotherapy resistance in ACC. Additionally, it provides a foundation for the development of more complex models incorporating stromal and immune cells. Abbreviations 3-dimensional adrenocortical carcinoma angiotensin II extracellular matrix forskolin high-performance liquid chromatography short tandem repeat Contributor Information Sarah Feely, Discipline of Pharmacology and Therapeutics, School of Medicine, University of Galway, Galway, H91 V4AY, Ireland. Nathan Mullen, Discipline of Pharmacology and Therapeutics, School of Medicine, University of Galway, Galway, H91 V4AY, Ireland. Padraig T Donlon, Discipline of Pharmacology and Therapeutics, School of Medicine, University of Galway, Galway, H91 V4AY, Ireland. Eileen Reidy, Discipline of Pharmacology and Therapeutics, School of Medicine, University of Galway, Galway, H91 V4AY, Ireland. Ritihaas Surya Challapalli, Discipline of Pharmacology and Therapeutics, School of Medicine, University of Galway, Galway, H91 V4AY, Ireland. Mariam Hassany, Discipline of Pharmacology and Therapeutics, School of Medicine, University of Galway, Galway, H91 V4AY, Ireland. Anna Sorushanova, Discipline of Pharmacology and Therapeutics, School of Medicine, University of Galway, Galway, H91 V4AY, Ireland. Eduardo Ribes Martinez, Discipline of Pharmacology and Therapeutics, School of Medicine, University of Galway, Galway, H91 V4AY, Ireland; Science Foundation Ireland (SFI) Research Centre for Research in Medical Devices (CURAM), Biomedical Science Building, University of Galway, Galway, H91 TK33, Ireland. Peter Owens, Centre for Microscopy and Imaging, Anatomy, School of Medicine, University of Galway, Galway, H91 TK33, Ireland. Anne Marie Quinn, Department of Anatomic Pathology, Galway University Hospital, Galway, H91 YR71, Ireland. Abhay Pandit, Science Foundation Ireland (SFI) Research Centre for Research in Medical Devices (CURAM), Biomedical Science Building, University of Galway, Galway, H91 TK33, Ireland. Brendan Harhen, Biological Mass Spectrometry Core Facility, University of Galway, Galway, H91 TK33, Ireland. David P Finn, Discipline of Pharmacology and Therapeutics, School of Medicine, University of Galway, Galway, H91 V4AY, Ireland; Science Foundation Ireland (SFI) Research Centre for Research in Medical Devices (CURAM), Biomedical Science Building, University of Galway, Galway, H91 TK33, Ireland. Constanze Hantel, Department of Endocrinology, Diabetology and Clinical Nutrition, University Hospital Zurich (USZ) and University of Zurich (UZH), 8091 Zurich, Switzerland. Martin O’Halloran, Science Foundation Ireland (SFI) Research Centre for Research in Medical Devices (CURAM), Biomedical Science Building, University of Galway, Galway, H91 TK33, Ireland; Translational Medical Device Laboratory, University of Galway, Galway, H91 V4AY, Ireland. Punit Prakash, Department of Electrical and Computer Engineering, Kansas State University, Manhattan, KS 66506, USA. Michael C Dennedy, Discipline of Pharmacology and Therapeutics, School of Medicine, University of Galway, Galway, H91 V4AY, Ireland; Science Foundation Ireland (SFI) Research Centre for Research in Medical Devices (CURAM), Biomedical Science Building, University of Galway, Galway, H91 TK33, Ireland. Funding This work was funded by National Institutes of Health (NIH) (R01EB028848) and Science Foundation Ireland (SFI) (20/US/3676) and (13/RC/2073 P2). C.H. receives funding from Swiss 3R Competence Centre. Author Contributions C.D. and S.F. created the experimental design. C.D. and S.F. wrote the paper. S.F. performed the experiments under the guidance and assistance of N.M., P.T.D., E.R., O.C., R.S.C., M.H., and M.C. E.R.M. aided in the RT-qPCR experiments. P.O. aided in the imaging using confocal microscopy. B.H. and D.F. aided with the running of Mass Spectrometry. A.M.Q. evaluated the histology slides. C.H. provided the MUC-1 cell line for these experiments. Disclosures No potential conflict of interest was reported by the authors. Data Availability Some or all datasets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request. References Associated Data Data Citations Data Availability Statement Some or all datasets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.