A Novel Methodology For Isolation and Primary Culture of Swine Gastric Epithelial Cells.

Background: Culture of primary epithelial cells has a great advantage over tumor-derived or immortal cells lines since functional phenotype and genetic makeup are preserved. Swine model has proved to be helpful and reliable as a surrogate model in human diseases. Several porcine cell lines have been established from a variety of tissues and shown to extensively contribute to the current understanding of several pathologies, including cancer. However, few protocols for the isolation and culture swine gastric epithelial cells with phenotype preservation have been described. Therefore, the objective of this research was to develop a new methodology for establishing a primary cell culture from the fundic gland area of the porcine stomach. Results: Enzymatic disaggregation of gastric tissue by using a combination of collagenase type I and dispase II, protease inhibitors (soybean trypsin inhibitor and bovine serum albumin), and antioxidants (Dithiothreitol) allowed the isolation of gastric epithelial cells from the fundic gland area with viability > 90% during the incubation period. Gastric epithelial cells cultured in RPMI 1640, DMEM HG, and DMEM/F12 media did not lead to cell adhesion, cluster formation and cell proliferation. By contrast, Williams’ medium supplemented with growth factors supports the conuence and proliferation of a pure epithelial cell monolayer after 10 days of incubation at 37 o C in a 5% CO 2 incubator. Mucin-producing cell phenotype of primary isolates was conrmed by PAS staining as well as the expression of MUC1 and MUC20 genes by RT-PCR and DNAc sequencing. Swine Gastric epithelial cells also showed origin-specic markers such as epithelial membrane antigen (EMA), cytokeratin cocktail (AE1/AE3) and cytokeratin 18 (CK-18) detected by immunohistochemistry and immunouorescence, respectively. Conclusions: A new methodology was successfully established for the isolation of primary gastric epithelial cells from the fundic gland area in a swine model, based on a combination of tissue specic proteases, protease inhibitors, and antioxidants. The formulation


Background
Gastric diseases represent a frequent cause of economic loss in the porcine industry around the world [1,5]. For example, ulceration of the non-glandular stomach is a common pathology in pigs, with a global prevalence of 93% [2]. The disease outcome involves several clinical symptoms as a consequence of gastrointestinal bleeding including anemia, weight loss, ulcer perforation involving peritonitis, and sudden death [2,5]. The pathogenesis of porcine gastric ulceration is not well known. It has been suggested that diet particle size, management strategies, genetic background, hormonal changes, gastric microbiota composition and infectious agents such as Helicobacter suis may be involved [3]. H. suis mainly colonizes the fundic gland area of the stomach, which produces persistent in ammation [2]. Several studies have concluded/suggested a strong association between the presence of H. suis in this glandular area and the prevalence and severity of lesions in the non-glandular stomach [2,6,7]. Therefore, the use of swine primary gastric cells derived from the fundic gland area may shed light on the biological process involved in the development of this multifactorial disease.
On the other hand, swine is considered a valuable animal model due to their physiological and anatomical similarities to humans [4,10]. For instance, the porcine model has been extensively used for the understanding of the pathophysiology of diabetes and coronary artery disease both associated with atherosclerosis and hypercholesteremia, considering their similarities in the cardiovascular anatomy and the pro les of lipid metabolism among both species [8]. This model also has a high level of DNA sequence homology and similar chromosome structure to humans. Therefore, new proteomic and genomic approaches to evaluate malignant tumors in the swine model are rapidly increasing [9][10][11][12].
Additionally, cancer biology research relies on relevant biological models to explain the mechanistic process of tumorigenesis in humans. Contrary to murine, the effect of human oncogenic mutations in cell growth, differentiation and apoptosis are similar in porcine primary cells. The latter group has been used in lymphoma and osteogenic tumor development under the induction of tumorigenesis in primary hematopoietic and broblastic cells, respectively. [9,13,14] Furthermore, swine primary cell lines have been established from different anatomic tissues such as mammary glands [11], kidney [12], intestine [15], trachea [12], lung [16], and alveoli [20]. These primary cells were used to evaluate gene expression patterns, drug susceptibility and cell physiology [22]. However, as for swine stomach, few protocols to isolate and culture gastric cells have been described, but there have not been improvements for phenotype conservation [23][24][25][26][27]. Therefore, the objective of this research was to develop a new methodology for establishing a primary cell culture derived from the fundic gland area of the porcine stomach. The methodology for new cell culture is appropriate to study not only human diseases but also swine gastric pathologies.

Animal sample collection
All animal procedures were approved by Ethics Committee for animal experimentation of Universidad de Antioquia (Approval Number 121/2018), conformed to the Colombian Regulations for Animal Use in Biomedical Research (Act 8430 of 1993 and 84 of 1989.). Fresh stomach tissues were obtained from three young adult male pigs (Sus scrofa domesticus) age of 30 -34 weeks and a mean weight of 80 kg (± 8 kg). These animal tissues were kindly donated by a private porcine farm owner at a local slaughterhouse called "Vijagual". Approval for pig research use was registered with the written inform consent from the farm owner. All the animals were in good body condition and considered disease free by the veterinary medical o cer, responsible for the health and hygiene of the slaughterhouse "Vijagual". From pig's stomach, 25 -100 grams of the fundic gland area were dissected and stored at 4 o C in 50 mL conical centrifuge tubes with transport medium, which consisted of DMEM HG (Catalog number:

GEC isolation and culture
Fundic glandular tissues taken from swine stomachs (n=3) were transferred into 100-mm cell culture plates containing 10 mL of fresh transport medium. Excessive fat, connective and muscular tissue was removed manually by dissecting the mucous membrane layer (epithelium) using sterile tweezers and surgical scissors. The epithelial layer of the mucous was peeled off by gentle scraping. Tissue was disrupted into pieces of approximately 1 mm 3 in size, later transferred into 50 mL conical tubes and centrifuged at 80 g for 10 min at 4°C. Supernatants were discarded and tissue fragments were resuspended in digestion medium and kept in agitation on a rotational shaker at 150 rpm for 2 hours at 37ºC. Thereafter, the resulting cell suspension was ltered by using sterile gauze and washed three times to eliminate mucous and centrifuged at 80 g for 10 min. Supernatant was removed carefully and GEC viability was evaluated with the trypan blue dye exclusion test (T8154-Sigma-Aldrich, St. Louis, USA). Cells were seeded at a density of 3.5x10 5 cells/well in proliferation medium onto 12-well plastic plates, previously treated with 400 μL bovine collagen type I coating solution (125-50 -Sigma-Aldrich, St. Louis, USA). Finally, cells were incubated for 24 hours at 37°C, under a 5% CO 2 humidi ed atmosphere. On the next day, non-adherent cells were removed by washing each well twice with pre-warmed HBSS. After this, fresh proliferation medium was immediately added and replaced every other day. On the eighth day of incubation, proliferation medium was modi ed by using half of the initial concentration of EGF (25 ng/mL), insulin (2 μg/mL), and FBSi (10%).

GEC growth rate and proliferation kinetics
Once GEC conditions for in vitro culture were optimized, 1x10 5 cells were seeded in triplicate in Williams' supplemented medium onto 12-well plastic plates previously treated with collagen type I and incubated for seven days at 37°C and 5% CO 2 . A growth curve was generated to identify the exponential and stationary phases by plotting out the number of viable cells. For this purpose, cells were collected every 24 hours, centrifuged and counted in a hemocytometer using trypan blue dye at 0.4% (T8154-Sigma-Aldrich, St. Louis, USA). Additionally, the proliferation kinetics was measured for up to 72 hours using WST-1 proliferation assay (5015944001-Sigma-Aldrich, St. Louis, USA). Brie y, 1x10 4 GEC were seeded onto 96-well plates, being previously treated and cultured as described above. Cells were harvested at 24, 48 and 72 hours, 10 µL of WST-1 reagent were added to each well and incubated at 37°C in 5% CO 2 for 2 hours. All samples were analyzed using the iMark Microplate Reader (Bio-Rad, Hercules, CA, USA) at 540 nm. These experiments were repeated twice under the same conditions.

Hematoxylin and eosin staining (H&E)
Epithelial morphology of GEC was evaluated by H&E staining. 1 x 10 4 cells were harvested from a seven-day monolayer and seeded onto microscope slides and incubated for 24 hours at 37°C in 5% CO 2 . The slides were dipped into a Coplin jar containing Mayer's hematoxylin dye for 30 seconds and rinsed twice with PBS during 1 min each; then, 1% eosin Y solution was added for 30 seconds. Images were taken using the Eclipse 2000 optical microscope (Nikon, Tokyo, Japan).

Mucin detection in GEC by Periodic Acid-Schiff (PAS) staining
GEC phenotype was con rmed by the Periodic Acid Schiff (PAS) staining in cell cultures collected on day 0, 7 and 15.
Brie y, 1 x 10 4 cells were harvested at each time point, seeded onto microscope slides previously covered with collagen type I and incubated for 24 hours at 37°C in 5% CO 2 . Then, GEC were xed with 4% paraformaldehyde dissolved in phosphate-buffered saline (PBS) for 15 min and rinsed with PBS. Slides were exposed for 5 min to 0.5% periodic acid solution and then stained with Schiff's reagent for 15 min. Schiff's reagent was removed and the slides were rinsed with running tap water for 10 min. Finally, cells were counterstained with hematoxylin solution for 5 min. All steps were performed at room temperature (RT). A homogeneous red-purple color inside the cytoplasm was considered positive for PAS staining. Images were taken using Eclipse 2000 optical microscope (Nikon, Tokyo, Japan).

MUC1 and MUC20 ampli cation by reverse transcription-polymerase chain reaction (RT-PCR)
Mucins 1 and 20 (MUC1, MUC20) expression in GEC was assessed by RT-PCR. Total cellular RNA was extracted from 1 x 10 5 GEC harvested on day 0, 7 and 15. Firstly, cells were washed twice with PBS and centrifuged at 80 g for 10 min at RT. Secondly, 1 mL of RiboZol RNA extraction reagent (VWR Life Science, Radnor, PE, USA) was added to 1.5 mL tubes and the extraction was performed according to the manufacturer's instruction. Thirdly, puri ed RNA was measured using

Results
Cell isolation and establishment of GEC culture The optimized enzymatic methodology allowed the isolation of GEC with high viability (>95%) after tissue digestion from the fundic gland area of the porcine stomach. Either, presence of mucus on the cell culture nor disparity in pH of culture medium was observed. Four different proliferation media were assessed to determine the best culturing conditions for GEC. For this purpose, cells cultured in DMEM HG, DMEM/F12, RPMI 1640 and Williams' media were evaluated in terms of cell viability and adherence (at least 20-30% of cell con uence during the rst 48 hours of incubation) and cell proliferation (80-100% of con uence 10 days after the seeding). The results showed that cells seeded in DMEM HG, DMEM/F12 and RPMI 1640 media did not allow the adherence and proliferation of GEC. Moreover, the microscopic inspection of cells revealed necrotic-like features such as nuclear condensation, loss of cell membrane integrity and ghost-like cells as well as abundant cellular debris (unpublished data). In contrast, Williams' medium facilitated the establishment of GEC, characterized by their initial adherence, followed by cell proliferation in clusters and absence of necrotic-like bodies ( Table   2).
On the other hand, GEC reached 20% and 100% of con uence in Williams' medium after three and ten days of seeding, respectively (Figure 1a and 1b). In addition, GEC stained with H&E showed a morphology corresponding to an epithelialcell origin such as a prominent nucleus, polygonal shapes and consistent size (Figure 1c). Additionally, GEC were tested for Mycoplasma spp contamination using the uorescent dye DAPI. Microscopic evaluation showed that the cytoplasm of stained GEC was free of nuclear bodies, which con rmed the absence of infection (Figure 1d). Finally, GEC could be passed in culture for up to two months, preserving their epithelial morphology and growth rate (unpublished data). In summary, the aforementioned results revealed that Williams' medium provided the best conditions for establishing and culturing GEC as primary cells from the fundic gland area of the porcine stomach.
GEC viability and growth rate WST-1 assay proved that cell viability remained the same during the rst 48 hours of culture. However, at 72 hours, optical density increased due to the proliferation of GEC. Such increase correlates with the results observed in the trypan blue dye experiments (Figure 2b), in which a two-fold increase in cell number was observed on day four (exponential phase); whereas on day six and seven, cell counting remained the same (stationary phase) with a monolayer 90-100% of con uence, producing contact-dependent inhibition of growth in primary culture (Figure 2a). Considering both cases, these data con rm that primary isolates of GEC from swine stomach preserve standard features of commercial immortal cell lines such as stable viability, logarithmic and stationary growth phases, and contact-dependent inhibition.

Mucins expression in GEC
Mucins are glycoproteins covering gastric epithelium highly expressed in the stomach [27]. In order to con rm the phenotype of GEC isolated from the fundic gland area of the porcine stomach, the detection of MUC1 and MUC20 genes by RT-PCR on days 0, 7 and 15 of culture was performed. The gene coding for ACTB was also ampli ed as an internal control. RNA aliquots at different times allowed the RT-PCR ampli cation of the gene MUC1 and MUC20. Same intensity bands were observed in the agarose gel, con rming the expression of mucins in a sustained manner, regardless of the time of culture. Ampli cation of the RNA internal control was also detected, suggesting that the presence of target genes was stable within the samples (Figure 3). On the other hand, MUC1 sequence analysis showed 100% homology with reference sequence XM_021089728.1 from Sus scrofa domesticus, revealing the speci city of the molecular biology assay ( Figure  6).
Additionally, to determine the expression pro le of neutral mucins, GEC were evaluated using an optical microscopy with PAS staining. Staining evaluation of PAS-treated GEC showed a purple-red color pattern in the cytoplasm, typical of mucin producing cells. Cells also presented an epithelial-like shape and regular nucleus size ( Figure 4a). These data allowed concluding that cell culture from swine stomach tissue are GEC since mucin genes and protein expression were detected in a constant manner. Moreover, Williams' medium enriched with speci c growing factors supported the expression and synthesis of these glycoproteins in vitro.

Expression of epithelial markers in GEC
To verify the nature of GEC, the expression of three different epithelial markers, namely, EMA, AE1/AE3 and CK-18 were evaluated by immunohistochemical and immuno uorescence techniques, respectively. EMA is a glycoprotein expressed in the apical membrane of most glandular and ductal epithelial cells. Cytokeratin is found as laments inside the cytoplasm and usually associated with the cytoskeleton of epithelial cells. In addition, AE1/AE3 detect several cytokeratins simultaneously but it does not detect CK18. The optic evaluation of GEC showed brown homogeneous staining mainly extended through the cell membrane, corresponding to the expression of EMA. At least 80% of GEC were positive for this epithelial marker (Figure 4b). Additionally, the expression of AE1/AE3 was con rmed by the presence of a grey staining distributed as laments in the cytoplasm with 80 to 90% of positivity of the target cells (Figure 4c). Fluorescent positive cells for CK-18 were also observed in 80-90% of the examined microscope elds with a cytoplasmic pattern. Additionally, nuclear counterstained with DAPI demonstrated the integrity and the epithelial shape of GEC ( Figure 5). These results con rmed that GEC enzymatically isolated from fundic gland area of the porcine stomach but maintained epithelial phenotype and mucin-expressing cell genotype

Discussion
The establishment of epithelial cell culture derived from fundic gland zone of porcine stomach tissue was proposed as a tool for studying different types of gastric disease in both human and swine. In this regard, several works have shown the usefulness and applicability of porcine models, rather than comparing to mice, as these can resemble tissue architecture and pathophysiology features in a better way. For example, research on cardiovascular diseases, organs transplantation, diabetes, wound repair and cancer among others, have been performed with wild-type pigs, providing conclusive data for translational research [14]. Moreover, porcine airway and intestinal epithelial cells were used for studying important emerging zoonotic diseases such as In uenza virus [30]. Also, gastric parietal cells were used to demonstrate how H. suis affects the viability and function of this cell type [23]. Accordingly, the establishment of GEC culture helps elucidate the biological process involved in the development of many diseases, including gastric conditions in humans and swine.
Several enzymatic protocols for isolating GEC have been described, which include enzymes such as collagenase type 1 [26,27,39] or in combination with pronase [24,25] and dispase [23]. However, previous methodologies do not include the use of protease inhibitors to neutralize the effects of proteolytic enzymes and antioxidants to preserve the cellular redox environment. Therefore, a new method for isolating epithelial gastric cells was evaluated, rst consisting of the mechanic and enzymatic disaggregation of the gastric tissue. The combination of collagenase type I, dispase II, protease inhibitors and antioxidant showed their best performance in terms of GEC viability and tissue dissociation during the incubation period. Collagenase type I is widely used in methodologies related to tissue dissociation and cell culture due to its protease potential, capable of digesting collagen bers present in connective tissue. On the other hand, collagenase type I has been used for broblast isolation [31] and culture of small intestinal epithelial cells [32]. Dispase II is a collagenase type IV with a mild proteolytic activity that has been used for isolation of labile primary cells such as porcine urothelial cells, lymphatic and embryonic endothelial cells, stem cells and Schwan cells [33][34][35][36]. In addition, protease inhibitors (STI and BSA) is recommended to protected cells from non-speci c proteolytic activity [37] and antioxidant (DTT) preserved in a cellular redox environment [38]. Therefore, the compounds present in the digestion medium were helping the preservation of the viability and cellular architecture of GEC.
On the other hand, the best performance observed in terms of cell culture was obtained using Williams' medium supplement with mitogens in comparison to RPMI 1640, DMEN HG and DMEM/F12. This was because of the induction and maintenance of cell adherence and proliferation, allowing cell growth for up to two months. Williams' medium was originally described for isolation and growing hepatocytes [40], and to best of our knowledge, it had not been used in isolation of GEC before. This medium includes higher glucose concentrations, essential amino acids, vitamins, among others, that allowed inhibiting apoptosis and supporting long-term cell culture. In addition, GEC had been previously isolated using well-described media such as Ham's F12 [39], DMEN [27], DMEM/F12 [23] and RPMI 1640 [25].
Moreover, cell culture medium must have mitogens to support the growth rate and proliferation of primary cells while maintaining the tissue-speci c characteristics [41]. In this study, Williams' medium was supplemented with mitogens such as EGF and insulin. EGF has been involved in a variety of physiological responses, including cell survival, proliferation and differentiation in several tissues. EGF has been described to have an anti-apoptotic effect in gastric epithelial cells infected by H. pylori [42] and to regulate the production of mucus as well as the repair of airway epithelium after injury [43]. Additionally, insulin helps cells to metabolize glucose and amino acids stimulating growth and proliferation of rabbit gastric epithelial cells [44]. In addition, there is evidence that the combination of insulin and EGF improves the proliferation of gastric epithelial cells [27]. Therefore, the isolation of GEC Williams' medium supplemented with mitogens (EGF and insulin) is proposed, improving cell viability and proliferation as well as preserving the phenotype, which makes this model reliable for experiments that need long-time observations. Cellular adhesion to the extracellular matrix during the establishment and maintenance of primary culture is a crucial step to preserve the viability and proliferation of these cells [41]. Several matrix coatings have been described to promote cell attachment in GEC, including a glass surface coated with bronectin or collagen derived from human placenta [24], Matrigel [23] and Vitrogen gel [39]. In this study, GEC was seeded in bovine collagen type I, which helps preserve ratio proliferation and culture for up to two months, preserving their epithelial morphology and cellular architecture.
We also con rmed that the isolated GEC preserved gastric markers. In this regard, the expression of the MUC1 and MUC20 genes was con rmed by RT-PCR. Mucins are glycoproteins produced by specialized epithelial cells located in the lumen of organs of the digestive, respiratory and reproductive tracts. They help in the preservation of the lubrication and integrity of the epithelial lining [29]. MUC1 and MUC20 are membrane-bound mucins constitutively present in epithelial cells that have been used as makers in GEC and alterations in their expression are related to H. pylori infections and gastric cancer [17][18][19]. Hence, the expression of MUC1 and MUC20 in GEC con rms that the gastric epithelial phenotype of these cells was preserved.
The expression of epithelial markers was also determined to con rm the epithelial phenotype of GEC by immunohistochemical EMA and the AE1/AE3, as well as immuno uorescence CK-18 detection. Positive staining was found in GEC for all markers. EMA was restricted to the GEC cell membrane surface and the pan-cytokeratin marker and CK-18 were distribute throughout the GEC. These markers are normally expressed in GEC, but their high expression means a characteristic of human cancer [21]. All these data con rm the functional phenotype of GEC, and hence, the need to establish a new method for the isolation of swine GEC.
Finally, it is well known that the use of human tissues, speci cally form solid organs, has clinical and ethical implications that preclude the availability of biopsies for basic research. Therefore, the use of porcine samples resembling the organ properties of humans offers several advantages, given the recent growing interest in non-rodent translational models for the study of human diseases. The protocol described here contributed to the development of translational research in acute and chronic gastric diseases in humans and swine.

Conclusions
A new method is herein reported, based on the evaluation of different cell culture media, from which Williams' medium supplemented with mitogens (FBSi, EGF and insulin) provides the best conditions for GEC proliferation and maintenance of functional features such as mucin production and expression of epithelial markers (EMA, CK AE1/AE3 and CK-18) in the long term. The methodology described in this paper for GEC isolation and culture allows the development of new in vitro applications for the study of physiological and pathological aspects in toxicology, pharmacology, microorganism interactions, and in ammation in human and swine gastric diseases.

Competing of interests
The authors of this paper declare that there is no con ict of nancial and non-nancial interests regarding the publication of this paper.  Tables   Table 1 List of primers, annealing temperature and amplicon size for Sus scrofa domesticus gene ampli cation used in the RT-PCR protocol.    Growth and cell viability of GEC cultured in William's medium. a. GEC cultures were evaluated daily during seven days with trypan blue for establishing cell growth and proliferation curve. b. GEC cultured in 96 well plates were exposed to WST-1 reagent for viability determination by optical density using an ELISA reader. Figure shows the mean SD of cell numbers and OD at 540 nm.  Expression of CK-18 marker in GEC. a. GEC were grown on collagen pre-coated sterile slides and incubated for 24 h, then xed with 4% paraformaldehyde, permeabilizated with a solution containing Triton X-100. GEC were analyzed under microscope in differential interference contrast (DIC). b. stained with DAPI dye for nuclei visualization. and c. labeled with an anti CK-18 antibody with the Alexa Fluor 555 Figure 6 Alignment of MUC1 sequences. RT-PCR was performed in order to verify the constitutive expression of mucins in GEC, like MUC1. Ampli ed product of 437 bp was extracted form agarose gel and puri ed for Sanger sequencing. Sequence was aligned with the swine muc1 reference annotated sequence (Accession number XM_021089728.1) by Clustal Wallis.
Background colored bases nucleic bases correspond to sequence homology.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. ARRIVEGuidelinesChecklistJuly2020Final.docx