Modeling Acute Ethanol-Induced Impairment with Gastric Organoids

Background: Ethanol have been linked to atrophic gastritis and gastric carcinoma. Although it is well known that ethanol can result in hypochlorhydria, the molecular mechanisms underlying this phenomenon remain poorly understood. Results: Here we used gastric organoids to show that ethanol permeabilized the apical membrane of gastric parietal cells and induced ezrin hypochlorhydria. The functional consequences of ethanol on parietal cell physiology were studied using organoids. Gastric organoids were pre-incubated in the basic medium or with EGTA or E64 , and incubated at 37 ℃ in either medium alone, or medium containing 6% ethanol. We assessed ezrin proteolysis. Ethanol permeabilization induced activation of calpain (cid:0) and subsequent proteolysis of ezrin, which resulted in the liberation of ezrin from the apical membrane of the parietal cells. Signicantly, expression of calpain-resistant ezrin restored the functional activity of parietal cells in the presence of ethanol. Conclusion: Taken together, our data indicated that ethanol disrupted the apical membrane-cytoskeletal interactions in gastric parietal cells and thereby caused hypochlorhydria. I in the hydrolysis of parietal cells Ezrin. Our results indicated that calpain I was present in gastric parietal cells and that ezrin was a substrate within the cell of calpain I, and the impairment by EtOH was demonstrated by protein immunoblotting from the level of the human organoids as well, which indicated that this mechanism may also applicable in human.


Background
The stomach tends to be exposed to various toxic factors, and ethanol is the most common factor that might contribute to the pathogenesis of various gastrointestinal (GI) dysfunctions, including gastritis, gastric ulcer, even gastric carcinoma 1 . Currently, the majority of knowledge about the cellular and molecular mechanisms of toxicity caused by ethanol relies on experiments performed in animal or cell models. However, the stomachs of different species vary widely in their morphology, the intricacy of their stomach networks and their topographical organizations. Therefore, it is di cult to gain a precise understanding of the effects of acute alcoholic-induced parietal injury in humans.
Gastric acid secretion and the process of regulation by the cAMP-dependent protein kinase, protein kinase A (PKA) pathway are involved in phosphorylation of the cytoskeletal protein ezrin, which was an 80-kDa microvilli-rich micro vessel. Ezrin protein belongs to the family of widely distributed cell membraneassociated proteins, the ERM protein family, whose initial isolation is based on the dependence of ezrin on the phosphorylation of PKA in stimulated parietal cells 2,3 . Ezrin protein is highly expressed in the stomach, lung, small intestine, and kidney, and also expressed in the spleen, thymus, bone marrow, and lymph nodes, but is extremely low in heart, brain, muscle, and testis 4 .It is mainly expressed in epithelial cells and distributed prior to the protrusion part of the cell surface: such as microvilli rich in actin, folds of cell membranes, lamentary pseudopodia rich in skeletal protein, stretched cell front, and various local adhesions and so on. Ezrin is polarized in the apical membrane of gastric parietal cells and is an essential component of gastric acid secretion 5 . When ezrin proteins were knocked out in mice, the vesicles of the proton pump to the apical membrane were unable to form and the normal functions were lost 6 , leading to severe defects in gastric acid secretion and death within three weeks 7 .
Elevated intracellular calcium concentration [Ca 2+ ] is the trigger for many cellular processes in response to external stimuli. Calpain, a calcium-dependent calpain found in the mammalian cytoplasm, plays a role in the regulation of intracellular signal transduction cascades of calcium ions. In many animal cells, the calcium-dependent proteolytic system consists of at least three proteins: the micromolar calcium demanding protease (calpain I), the millimolar calcium demanding protease (calpain II), and natural endogenous inhibitor, calpain inhibitor which is the speci c inhibitor that inhibits calpain I and II. Calpain is critical for the development of pathology and is a unique and powerful therapeutic target.
Human adult stomach stem cells have been recently reported to enable the self-organization of millimeter-size, complex 3D stomach tissues with unprecedented morphogenetic and histological realism in vitro, referred to as gastric organoids [8][9] . This near-physiological and self-renewing 3D organoid sustain similar organ functionality as the stomach. Although reductionist in nature, organoids have great potentials to bridge between traditional 2D culture and animal models, and open up new avenues for stomach study and disease modeling. To date, gastric organoids have been used to establish various models of gastric diseases including H. pillory infection 10 and proved to be powerful as a 3D platform for the investigation of psychiatric disease origin and pathology in vitro. Herein, we presented a stem cellbased gastric organoid model that allows the exploration of acute ethanol-induced pathogenesis of gastric development in vitro (Fig. 1).When exposed to ethanol, gastric organoids were signi cantly impaired in comparison with the controls, including the diameter and morphology, and as ezrin proteolysis.
To observe the changes of protein markers in the growth process of gastric organoids, we xed the murine gastric organoids on day 1, 3, 5 and 7 respectively, and H + , K + -ATPase, ezrin, F-actin and DAPI of parietal cells were stained, and photos were taken with confocal microscope (Fig. 1). Ezrin protein, a marker of the parietal membrane, indicated that the parietal membrane of murine gastric organoids was oriented towards the lumen and had polar distribution. The presence of H + , K + -ATPase, a marker of parietal cells, suggested that the model system could be used to study the physiological and pathological characteristics of parietal cells.
In order to further semi quantitatively observe the changes of protein markers in the growth process of stomach organoids, murine gastric organoids were established, and collected on the 1 st , 3 rd , 5 th , 7 th , 9 th and 11 th day respectively, and the distribution of H + , K + -ATPase, ezrin and F-actin markers in the parietal cell were detected. The results showed that the protein markers of murine gastric organoids remained stable in Day 5-9 ( Fig. 1, d).
In order to observe the effect of different concentrations of ethanol on gastric organoids, nal concentrations of 0%, 2%, 4%, 6%, and 8% EtOH were added to the culture medium to impair murine gastric organoids. One hour later, we stained them with 0.4% trypan blue solution and took photos under inverted uorescence microscope (Olympus X81). Statistical analysis of organoids with positive trypan blue staining was carried out. As shown in Figure 2, the nal concentration of EtOH was 6%. Trypan blue staining was found in 50% of the gastric organoids one hour after injury. However, when the concentration of EtOH exceeded 10%, the Matrigel was easy to dissolve and collapse, which affected the next observation and experiment.
In order to dynamically observe the impairment of 6% ethanol on the gastric organoids, the living gastric organoids were stained with Hoechst for 30 minutes, and then were placed under the microscope. Ethanol with the nal concentration of 6% was added and the images were taken at the same time. The imaging time was 30 minutes. 30 minutes later, the nuclei of murine gastric organoids shrank and the lumen became narrower (Fig. 2, b).
Considering that the center of the gastric tube was a cystic luminal structure, and due to the concentration limitation of ethanol added to the medium, microinjection was used to inject the ethanol into the lumen of the gastric organoids (Fig. 2, c). One hour later. The samples were subjected to western blotting or immuno uorescence staining to observe impairment on gastric organoids. However, one hour after 6% ethanol was injected into the gastric organoids, no obvious impairment was observed. Next, we increased the concentration of ethanol to 50%. After 1 hour, we observed the death of organoids. As shown in Fig. 2, c, the gastric organoids were cultured, and microinjections of ethanol at different concentrations were performed on Day 6 of the organoids. One hour after microinjection of 50% ethanol into the organoids, the apical membrane marker ezrin and the parietal cell marker H + , K + -ATPase were destroyed, and the lumens of the organoids collapsed.
Next, the murine gastric organoids were impaired with EtOH added to the medium at a nal concentration of 6%, and western blotting was performed at 0,15,30,45, and 60 minutes. It was found that there was a cleaved ezrin in 55kDa and the band had a tendency of becoming more and more obvious over time (Fig.2, d).
Immuno uorescence staining of ezrin, F-actin and DAPI (Fig. 3, a), or H + -K + -ATPase, ezrin and DAPI ( Fig.  3, b) was performed 1 hour after gastric organoids were impaired by 6% ethanol added to the medium with a time gradient of 0-90 min. As can be seen from Fig. 3, a, with the prolongation of time, the location of the parietal membrane of ezrin gradually became diffuse, and the lumen of organoids shrunk.
Ethanol-induced organoids impairment was caused by calcium-dependent calpain activation.
To investigate whether the impairment of ethanol-mediated gastric organoids was caused by calciumdependent calpain activation, gastric organoids were divided into four groups: 1.8 mM CaCl 2 group, 1.8 mM CaCl 2 + 6% ethanol group, 1.8mM CaCl 2 +6% ethanol + 100μM E64 group and 30μM BAPTA+1.8mM CaCl 2 +6% ethanol group. The murine gastric organoids were injured for 1 hour, and then samples were taken to run western blot to detect calpain I, ezrin and F-actin. Image J was used to measure the gray value of the strips. As a result, the hydrolyzed band of ezrin in the 1.8 mM CaCl 2 + 6% ethanol group increased, and the difference was statistically signi cant(p<0.001) (Fig. 3, d).
Effects of ethanol-mediated calcium ion impairment on the localization of ezrin, H+, K+-ATPase and Factin in murine gastric organoids.
Over time, the localization of ezrin in the apical membrane of organoids in the 1.8mM Ca 2+ +6% EtOH group became blurred and diffused, while the localization of ezrin in the groups of 1.8mM Ca 2+ +6% EtOH + EGTA group, 1.8mM Ca 2+ +6% EtOH + BAPTA group and the 1.8 mM Ca 2+ + 6% EtOH + E64d group remained mostly in the apical membrane and was clear (Fig. 4).
To investigate the effect of ethanol-mediated calcium ion on ezrin expression in human gastric organoids, we injured human gastric organoids either with 1.8 mM CaCl 2 , 1.8 mM CaCl 2 +6% EtOH, or with 1.8 mM CaCl 2 +6% EtOH + E64, respectively. After 1 hour, samples were collected for western blot, and the expressions of ezrin and actin were detected, and the values of the bands were measured with Image J (Fig. 5).
As can be seen from Fig. 5, ezrin in the 1.8 mM CaCl 2 + 6% EtOH group showed signi cant hydrolysis, suggesting an increase in the activity of calpain I. In the E64d group, no signi cant hydrolysis of ezrin was observed. There was no signi cant increase in calpain I activity, and the difference was statistically signi cant in the E64 group compared with the alcohol group (p < 0.05).

Discussion
Before the origin of organoid system, the studies of gastric diseases were mostly modeled using cell lines or animal models or explants, but these model systems have their own limitations 11 .
The gastric epithelial cell line is a commonly used model system for studying gastric diseases, which was originally derived from gastric tumors. Their advantages are immortalized, readily available, easy to culture and can be applied to a variety of experiments. However, its molecular characteristics have been modi ed, including tumor suppressor genes, oncogenes, and cell cycle regulatory genes, and thus there is a certain gap with epithelial cells in physiological state. Cell lines accumulate mutations during in vitro culture for decades, and some cell lines are infected with viruses, such as AGS cell lines, which are usually infected with type parain uenza virus, affecting pathways involved in immunity, proliferation and tumorigenesis, and interferon responses 12 .
Conventional tumor cell culture is two-dimensional, but studies have shown that whether cell lines grew in 2D or 3D can affect drug e ciency 13 . This suggests that 3D is an important feature of tissue response.
However, 3D cell spheres cultured from a single cell line lack a combination of different cell types typical of tissues and organs, while gastric organoids may include parietal cells, chief cells, cervical mucus cells, stem cells, and endocrine cells, etc.
Primary cell culture models are often a good alternative to tumor cell lines because they have not been transformed. Primary cultured gastric parietal cells have polarized structural features and polarized exocytosis and endocytosis, and were easy to access to large amounts. However, these cells are unable to self-renew and cannot be cultured and expanded for a long time, and all of the original cells need to be freshly isolated for each experiment. The animal model system is more holistic and dynamic than the cell line, but some human stomach diseases cannot be responded to by animal models. For example, after mice were infected with Helicobacter pylori, they often only have mild gastritis and would not progress to gastric ulcer or stomach cancer. Mongolian gerbils can progress to gastric cancer after infection with Helicobacter pylori, but its distant line limits its researches and applications 14 .

The gastric organoids include parietal cells, chief cells, cervical mucus cells, stem cells, and endocrine
cells which are also contained in the fundic glands. The parietal cells secrete hydrochloric acid and internal factors; the chief cells secrete pepsinogen; the enterochroma n-like (ECL) cells in endocrine cells secrete histamine, G cells secrete gastrin, D cells secrete gastrin, and cervical mucus cells Secrete mucus.
The organoids can last for more than a year without losing the ability to expand or differentiate 15 .
In addition, the three-dimensionally cultured gastric organoids have complex 3D structures with structures and functions similar to those of internal organs 8,16 . The gastric organoids are mostly cystic, and some will gradually grow a bud-like structure around the cyst. There is no difference in the culture time, the expansion ratio, the expression of the gene markers of each constituent cell, or the cell type in the gastric organoids derived from a single gastric stem cell and from the gastric gland 17 .
Organoids can also be cultured in two dimensions, and electron microscopic observation reveals that the apical membrane of the stomach of the two-dimensional culture is oriented toward the medium 18 . Since the gastric epithelium is a special polar epithelial cell, we found that the apical membrane of the gastric organoid faces the inner cavity by staining the apical membrane marker ezrin, and the polarity distribution is in consistency with the gastric mucosal epithelial cells.
The gastric organoids both replicate the complexity of the disease phenotype in vivo while retaining the accessibility of in vitro. Gastric organoids that can be cultured in vitro for a long time have enabled researchers to overcome many obstacles and conduct basal and translational studies 19,20 . Because it is closer to the environment in vivo, it is more faithful to the various physiology and pathology. Although organoids have their own limitations, bioengineering methods such as time-dependent and spatialtemporal materials can control the growth of organoids toward the desired structural form and selforganization, and other studies are already available. Vascular networks are added to organoid cultures to increase nutrients in a nearly physiological way 21,22 . Co-culture methods for organoids can also mimic multiple organ lesions 23 .
In this study, after different concentrations of EtOH were added to murine gastric organoids for 1 hour and after trypan blue staining, murine gastric organoids showed different degrees of positive trypan blue staining, indicating the impairment caused by EtOH. The extent of the damage of the gastric organoids depended on the concentration. Ethanol increases the permeability of parietal cell membranes 24 and triggers an increase in Ca 2+ . Ethanol and endogenous aldehydes impaired chromosomal and variant stem cells 25 .
Clinically, a variety of diseases and gastric acid secretion abnormalities are related, including gastroesophageal re ux disease, chronic gastritis, gastric ulcer, benign and malignant tumors of the gastrointestinal tract, anemia and so on.
In vivo methods for detecting gastric acid secretion in animal models often require an invasive procedure after anesthesia is xed in the animal, and the secreted gastric acid is collected after catheterization into the stomach and is subsequently measured 26 . The in vitro gastric acid detection method is more commonly used in the 14 C aminopyrine experiment which was rst proposed by Berglindh and Obrink in 1976 27 . Due to the investigator's radioisotope experimental quali cations, the 14 C aminopyrine test was temporarily unable to detect gastric acid.
Despite this, we observed that the number and location of most H + , K + -ATPase were not signi cantly changed in western blotting and immuno uorescence experiments, but in repeated experiments, we observed H + , K + -ATPase is activated in the control group and diffused in the alcohol group(data not shown), suggesting that alcohol inhibition of gastric acid secretion may be related to the activation of H + , K + -ATPase, possibly due to its short spontaneous activation time. A statistically signi cant number of activated H + , K + -ATPases are to be further explored in future experiments using H + , K + -ATPase lentivirusinfected organoids.
From the results of western blot and immuno uorescence, we found that ethanol mediated the hydrolysis of ezrin, which was indicated by a speci c hydrolysis zone at 55kD.After being impaired by 6% EtOH with different time intervals, a similar trend also emerged. The wrapping of Matrigel may have a certain physical protection effect on gastric organoids.
In this study, two methods were used to establish the ethanol-impairment models. The rst one was to add ethanol to the medium to simulate the effect of alcohol concentration in the blood on murine gastric organoids. The advantage of this method was that it was more feasible. The disadvantage was that when the ethanol concentration is greater than or equal to 10%, the Matrigel collapsed.
In the view of the fact that the gastric organoids are similar to the stomach and are hollow cystic structures of the lumen, we have explored the use of microinjection techniques to directly inject ethanol into the internal gastric organoids to simulate gastric gavage. It was found that this modeling method could avoid the in uence of Matrigel on the drug, and can be used to study the direct interaction of alcohol or pathogen-host, and at the same time, due to the surface tension of murine gastric organoids and the tight connection between cells, when the concentration of EtOH reached up to 50%, the Matrigel remained intact. The downside was that when 6% alcohol is injected into the cavity, no obvious impairment was observed, and it may be related to the dilution of the ethanol concentration with a small amount of liquid in the cavity. Because the intracavitary volume of the stomach was temporarily unable to be accurately estimated, it was di cult to determine the speci c nal concentration of ethanol injected into the cavity. In addition, it took a certain time to microinject the organs of each stomach, and it is inconvenient to carry out protein immunization after drug stimulation. So, we chose the rst modeling method that was more suitable for our experimental purposes. BAPTA-AM is a cell permeable chelating agent with high selectivity for Mg 2+ and Ca 2+ , which can be used to control intracellular Ca 2+ levels and chelate intracellular calcium ions. BAPTA has a higher selectivity for Ca 2+ than EDTA and EGTA, and its metal binding is also less sensitive to change of pH result. In this experiment, BAPTA chelated intracellular calcium ions, and the decrease of intracellular calcium ions caused the hydrolysis of ezrin to be signi cantly reduced compared with the alcohol group. EGTA chelated extracellular calcium ions, which reduced the in ux of calcium ions, and decreased the intracellular calcium ion, so hydrolysis of ezrin was signi cantly reduced compared with the alcohol group.
E64d (molecular formula C17H30N2O5, relative molecular mass 342.4) is a speci c calcium-activated neutral cysteine-based endopeptidase inhibitor with good cell membrane permeability and high selectivity for cysteine proteases. It can form a covalent bond with the cysteine thiol group at the active site of calpain, inhibiting its hydrolysis ability, and has a protective effect on alcohol-mediated gastric mucosal epithelial damage. In the E64d group, the hydrolysis of ezrin was signi cantly reduced, indicating that inhibition of calpain I activity was a factor determining the hydrolysis of ezrin, and demonstrated the role of calpain I-ezrin interaction in the mechanism of alcohol-mediated impairment of gastric organoids.
Calpain digests ezrin on parietal cells and impaired gastric acid secretion 24,28 . Ezrin in parietal cells binds to the apical membrane during resting and stimulating phases 29,30 , while some studies have suggested that ezrin was released from the apical membrane into the cytosol at rest in parietal cells 31 .

Conclusion
In summary, hydrolysis of gastric ezrin re ects Ca 2+ -dependent protease activity. The in-situ hydrolysis of ezrin was blocked by the transmembrane cytoplasmic protease inhibitor E64d, which enhanced the important role of calpain I in the hydrolysis of parietal cells Ezrin. Our results indicated that calpain I was present in gastric parietal cells and that ezrin was a substrate within the cell of calpain I, and the impairment by EtOH was demonstrated by protein immunoblotting from the level of the human organoids as well, which indicated that this mechanism may also applicable in human.

Ethics Statement
Mouse experiments were performed in accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and following the International Guiding Principles for Biomedical Research Involving Animals. Protocols were approved by the Institutional Animal Care and Use Committee of Beijing University of Chinese Medicine.

Source of animals
All animals were obtained from Vital River Laboratory Animal Technology Co., Ltd.

Establishment of gastric organoids
Mice were euthanized by CO 2 asphyxiation. Cultures were kept at 37°C in a 5% CO 2 incubator and the medium were refreshed every 2 days. Organoids were passaged every 6-7 days at a ratio of about 1:4. Statistical signi cance was de ned as p<0.05. All statistical analyses were calculated using GraphPad Prism (Version 5.2). All experiments were performed with at least 3 biological replicates (n≥3), each analyzed in at least 2 technical replicates.

Trypan blue staining
For vital staining, gastric organoids were incubated with trypan blue for 2-5 min, washed 3 times with PBS, left in 37℃for another 15 min with PBS, washed again and visualized.

Statistical analysis
The statistical analysis was performed using the statistical package for the social sciences (SPSS, version 13.0, Chicago, IL). The non-normally distributed data were expressed as the median (quartile range). The within group differences were assessed using a non-parametric test while the Wilcoxon rank test was used to compare two groups. The normally distributed data and homogeneous variances were expressed as the mean±standard deviation (SD). Multiple comparisons were performed using a one-way analysis of variance (ANOVA) followed by the least signi cant difference (LSD) test. A p value <0.05 was considered statistically signi cant for the analyses.  Ethanol impaired gastric organoids. a Trypan blue staining results of gastric organoids injured by EtOH with different concentrations (Bars= 200 μm). b Statistics of live cell photographs of gastric organoids impaired by EtOH. **p< 0.01.c Immuno uorescent images of gastric organoids at 0 min and 60 min after being injured by 50% EtOH by microinjection. d Western blot results of 6% EtOH-induced impairment to murine gastric organoids. **p <0.01. Ethanol-induced gastric organoids impairment was caused by calcium-dependent calpain activation. a Immuno uorescence of ezrin and F-actin in murine gastric organoids impaired by time gradient of 6% ethanol (bars= 50 μm).b Immuno uorescence of ezrin and H+, K+-ATPase in murine gastric organoids with 6% ethanol by time gradient (bars = 50 μm).c Western blotting and statistical analysis of murine gastric organoids of CaCl2 group, 6% ethanol group, E64 group and BAPTA group. d Western blotting and statistical analysis of murine gastric organoids of CaCl2 group, 6% ethanol group, EGTA group and BAPTA group. ***p<0.001.