A method for establishing a 3D reconstructed dry eye model induced by potassium sorbate under low humidity conditions

Abstract

The present invention belongs to the field of tissue engineering technology, and specifically relates to a method for establishing a 3D reconstructed dry eye model induced by potassium sorbate under low humidity conditions, and the construction method includes: screening of potassium sorbate induction concentration; construction of an in vitro reconstructed 3D corneal epithelial model; induction construction of an in vitro reconstructed 3D dry eye model by administering drugs on the surface of the model under low humidity conditions; and verification of the effectiveness and stability of the in vitro reconstructed 3D dry eye model. The method effectively simulates human dry eye symptoms and environment, and can also intuitively characterize the state of the dry eye model through inflammatory factors, tissue vitality, and histological structure. The dry eye model induced by the method is simple to operate, the model induction process is time-consuming, the cost is relatively low, the obtained model has good stability, and it is easier to achieve systematization and standardization. Similarly, this model uses human corneal epithelial cells to avoid species differences, and provides a new perspective in evaluating drugs or treatment methods, and in-depth research on the pathogenesis of dry eye and potential treatment strategies.

Description

Technical Field

[0001] The invention belongs to the technical field of tissue engineering, and particularly relates to a method for establishing a 3D reconstructed dry eye model induced by potassium sorbate under low humidity conditions. Background Art

[0002] Mixed dry eye is the most common dry eye in clinical practice, which is caused by two or more causes. The prevalence of dry eye is increasing year by year and has a negative impact on people's physical and mental health, work and life, and the development of social production and economy. Tear film instability, inflammation and damage are the main mechanisms leading to physiological and pathological processes. Although dry eye is common, there are few effective therapeutic drugs and treatment methods, which is related to the diversity of dry eye and the lack of dry eye models.

[0003] In addition, the choice of dry eye modeling method has a great impact on drug screening, and different modeling methods cause different dry eye diseases and severity. At present, corneal epithelial cells are the seed cells for constructing in vitro corneal epithelial cell models of dry eyes. However, 2D cell culture cannot completely simulate the microenvironment of cells in vivo, resulting in obvious differences in cell morphology, differentiation, cell-matrix interaction, and cell-cell interaction with the behavior of cells under physiological conditions in vivo, resulting in differences between experimental results and actual conditions in vivo. Animal models of dry eye are commonly used models for studying the pathogenesis of dry eye and exploring treatment methods. However, the establishment of animal models of dry eye is complicated, the method is difficult to operate, and requires high technical conditions and superb surgical skills or a long time. The activity of the cornea decreases after ex vivo, and there are too many uncontrollable factors and large variability. On the one hand, the use of different animal models has certain limitations, which is somewhat different from the actual clinical situation; on the other hand, surgery will cause greater irreversible trauma to animals, which seriously violates the spirit of the 3R principle. Although animal models of dry eye simulate some pathological features of dry eye in patients, no animal model can accurately reproduce all the symptoms and signs of human dry eye, especially the diversity of pathological changes and the chronic progressive characteristics of the disease. In summary, both 2D cell dry eye models and animal dry eye models limit their applications in disease modeling, tissue engineering, drug screening, and personalized medicine.

[0004] Potassium sorbate is an antibacterial preservative that is continuously being developed and applied in the food, beverage, skin care, pharmaceutical and other industries. Potassium sorbate is added to most eye drops. It is normal for some eye drops to have appropriate preservatives, but long-term and frequent use of eye drops containing preservatives may further aggravate dry eye, which can destroy the tight connection between the superficial surface cells of the corneal epithelial cell layer, causing tear film stability destruction, ocular surface inflammation, corneal epithelial cell apoptosis and damage, etc., and even cause drug-induced dry eye, the clinical manifestations and related tissue pathological changes are similar to human dry eye.

[0005] Many studies have found that a dry and low-humidity environment is one of the main reasons for rapid tear evaporation, shortened tear film breakup time, increased tear film osmotic pressure, and upregulated expression of ocular surface inflammatory factors. Modern medicine believes that a dry environment can cause rapid tear evaporation on the ocular surface, leading to dry eyes. Summary of the invention

[0006] The purpose of the present invention is to overcome the limitations of 2D cell and animal models in the fields of dry eye disease modeling, drug screening and personalized medicine, and establish a 3D reconstructed dry eye alternative model method induced by potassium sorbate under low humidity conditions, and study the stability of the model induced by this method, observe the stability of the 3D reconstructed dry eye model induced by local use of potassium sorbate under low humidity conditions, and determine the effective time of the model. This method can effectively simulate human dry eye symptoms, and can also intuitively characterize the state of the dry eye model through inflammatory factors, tissue vitality, and histological structure. In addition, the dry eye model induced by this method is simple to operate, the model induction process is time-consuming, the cost is relatively low, the obtained model is stable, and it is easier to achieve systematization and standardization. Similarly, this model uses human corneal epithelial cells to avoid species differences, and provides a new perspective in evaluating drugs or treatment methods, and in-depth research on the pathogenesis of dry eye and potential treatment strategies.

[0007] The present invention is achieved through the following technical solutions:

[0008] (1) Screening of potassium sorbate induction concentration;

[0009] Corneal epithelial cells were collected and cultured at 0.5 x 10 4 pcs / hole~1 x 10 4 The cells were inoculated into a 96-well plate at a density of 100 / well. When the cell fusion rate reached 70%-80%, the culture medium containing the corresponding concentration of sample was replaced. After 24 h, the cells were incubated in the dark for 3 h with MTT working solution. Finally, 100-200 μL of isopropanol was added to each well and the absorbance at 570 nm was read on a microplate reader.

[0010] (2) Construction of in vitro reconstructed 3D corneal epithelial model;

[0011] Corneal epithelial cells were collected and cells of passages P3 to P7 were used. They were seeded into Transwell chambers coated with Coating Matrix at a seeding density of 100 ± 50 w / well. The chambers were placed in 6-well plates for culture. The corresponding volumes of culture medium I were added inside and outside the chambers, with the liquid levels at the same level. On the 5th day of culture, culture medium II was replaced to start air-liquid interface culture, and the whole process of culture was 10-15 days.

[0012] (3) Inducing an in vitro reconstructed 3D dry eye model under low humidity conditions by simulating human administration;

[0013] The low humidity culture conditions are 30-40°C, 5% CO2, and RH 30-60%.

[0014] 1) Drug preparation: 0.05% to 1% potassium sorbate, the negative control group was given 15 μL to 50 μL PBS, and the experimental group was given 15 μL to 50 μL potassium sorbate;

[0015] 2) Model preparation: Add 0.9 mL of culture medium II to a 6-well plate, divide the experimental groups, and place one in vitro reconstructed 3D corneal epithelial model obtained in (2) in each well;

[0016] 3) Model drug administration: All drug administration operations were performed in a clean bench. On the 7th to 9th day of gas-liquid culture, drug administration was simulated in the human body. Drugs were drawn up with a pipette and slowly dripped onto the surface of the model. Drugs were administered to one model at intervals of 30 s to 2 min. The model was gently shaken to ensure that the drugs were evenly distributed on the surface of the model.

[0017] 4) Induction time: Administer once a day for 24 to 48 hours.

[0018] 5) Successful verification of dry eye model

[0019] ① Inflammatory factor detection: From the dry eye models obtained in 1) to 4), 3 samples were randomly selected from each group. After the administration time of the drug-administered tissues was over, the culture was collected and stored in EP tubes at -80°C for the detection of inflammatory cytokine TNF-α. GraphPad Prism Program software was used for drawing. t -test statistical analysis,* p <0.05 indicates significant difference, ** p <0.01 indicates extremely significant difference.

[0020] ② Tissue viability detection: Take out the model and start the washing procedure. Control the washing of each culture chamber for 30 seconds, wash the model 15 times with duplex buffer solution, and wipe it with sterile gauze or sterile cotton swab. Transfer to a 24-well plate containing 300μL MTT working solution, incubate in the dark for 3 hours, use a pipette to remove the MTT working solution, and completely immerse the dry eye model in a 24-well plate containing 1.5mL of isopropanol in each well. Shake on a shaker for 2 hours to obtain the isopropanol extract, and aspirate 200μL of the isopropanol extract into a 96-well plate and read the absorbance at 570 nm on a microplate reader. GraphPad Prism Program software was used for drawing, and the groups were compared using t -test statistical analysis,* p <0.05 indicates significant difference, ** p <0.01 indicates extremely significant difference.

[0021] ③ Histological structure detection: 4% paraformaldehyde was used for fixation and sent to a third-party testing agency to detect the histological structure of the model. Compared with the control group CK, the arrangement of each layer of the histological structure was still irregular, the tissue structure was not tight, and vacuoles appeared.

[0022] (4) Verification of the effectiveness and stability of the in vitro reconstructed 3D dry eye model;

[0023] ① Randomly select 3 dry eye models from each group obtained in (3). After the administration time of the drug administration tissue is over, take out the model and start the washing procedure. Control the washing of each culture chamber for 30 seconds, wash the model 15 times with duplex buffer solution, and wipe it with sterile gauze or sterile cotton swab.

[0024] ② After washing, transfer to a 6-well plate containing 0.9 mL of culture medium II and incubate for 0 h, 24 h, and 48 h, respectively.

[0025] ③After the incubation, the obtained dry eye model was subjected to corresponding testing using the same testing methods as above (testing histological structure, inflammatory factors, and tissue vitality).

[0026] Preferably, culture medium I comprises: the basic component is KSFM, and the added components are: 1-10% fetal bovine serum, 1-20 μg / mL insulin, 0.5-2% penicillin-streptomycin, and 0.5-2% L-glutamine, and the percentages are volume percentages.

[0027] Preferably, culture medium II comprises: the basic component is KSFM, and the added components are: 0.1-5% bovine pituitary extract, 1-20 μg / mL insulin, 1-10 ng / mL hydrocortisone, 0.5-10 ng / mL epidermal growth factor, 0.5-10 ng / mL basic fibroblast growth factor, 1-10 μg / mL transferrin, and 0.5-2% penicillin-streptomycin, and the percentages are volume percentages.

[0028] The beneficial effects of the present invention are:

[0029] (1) The present invention establishes a dry eye replacement model whose tissue structure and physiological function are very similar to those of dry eye corneal tissue.

[0030] (2) The present invention used the potassium sorbate-induced injury model for the first time, observed the stability of the 3D reconstructed dry eye model induced by local use of potassium sorbate combined with temperature and humidity control, and determined the effective time of the model.

[0031] (3) Use in vitro 3D recombination methods to construct models, replacing animal models and 2D cell models, ensuring animal welfare and making up for the shortcomings of 2D cells. This can avoid species differences and provide a new perspective in evaluating drugs or treatments, and in-depth research on the pathogenesis of dry eye disease and potential treatment strategies, thereby meeting the market demand for stable dry eye models.

[0032] (4) Potassium sorbate can easily cause damage to the ocular surface when administered on the human body surface. The modeling time of the 3D reconstructed dry eye model is shorter than that of the animal model, so it is more suitable for the establishment of a model of severe dry eye in the short term.

[0033] (5) The present invention controls the temperature and humidity in the incubator, which can better simulate dry eye syndrome in a real dry and low-humidity environment, thereby improving the accuracy of the dry eye experimental model.

[0034] (6) The dry eye model induced by this method is simple to operate, the model induction process is time-saving, the cost is relatively low, the obtained model is stable, and it is easier to achieve systematization and standardization.

[0035] After a large number of experiments, the present invention screened out that a dry eye model can be obtained by inducing 30μL0.5% potassium sorbate for 24h in a culture environment of 40°C, 5% CO2, RH30-60%; after administration and rinsing, incubated in a 6-well plate containing 0.9mL of culture medium II for 48h, the in vitro reconstructed 3D dry eye model still maintains its effectiveness, indicating that the model induced by the above method has good stability. The in vitro reconstructed 3D dry eye model obtained has increased inflammatory factors, reduced tissue vitality, partial loss of matrix layer cells and loose arrangement, reduced number of cell layers, and the appearance of vacuoles. Based on this, the above conditions are determined as the preferred conditions of the present invention after comprehensive consideration. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 The histological structure of the in vitro reconstructed corneal epithelium / dry eye model obtained under different conditions is shown;

[0037] Figure 2 The results of the inflammatory factor TNF-α detection in the in vitro reconstructed corneal epithelium / dry eye model obtained under different conditions are shown;

[0038] Figure 3 The results of testing the barrier indicators of the in vitro reconstructed corneal epithelium / dry eye model obtained under different conditions are shown;

[0039] Figure 4 The histological structure of the in vitro reconstructed dry eye model obtained after potassium sorbate induction administration and incubation for 24 and 48 hours respectively is shown;

[0040] Figure 5The results of the inflammatory factor TNF-α detection in the in vitro reconstructed dry eye model obtained after incubation for 0h, 24h, and 48h after potassium sorbate induction administration are shown;

[0041] Figure 6 The results of the barrier index detection of the in vitro reconstructed dry eye model obtained after incubation for 0h, 24h, and 48h respectively after potassium sorbate induced administration are shown. DETAILED DESCRIPTION

[0042] In order to enable those skilled in the art to better understand the scheme of the present invention and to make the above-mentioned objects, features and advantages of the present invention more obvious and understandable, the present invention is further described in detail below in conjunction with embodiments.

[0043] Example 1

[0044] (1) Screening of potassium sorbate induction concentration;

[0045] In order to successfully construct the model, different concentrations of potassium sorbate were first screened for induction concentrations.

[0046] Corneal epithelial cells were collected and cultured at 1 x 10 4 The cells were inoculated into a 96-well plate at a density of 100 cells / well. When the cell fusion rate reached 80%, the culture medium containing 0, 0.25%, 0.3%, 0.5%, 0.6%, and 1% potassium sorbate was replaced. After 24 h, the cells were incubated in the dark with MTT working solution for 3 h. Finally, 200 μL of isopropanol was added to each well and the absorbance at 570 nm was read on a microplate reader.

[0047] Table 1: Cytotoxicity of different concentrations of potassium sorbate

[0048]

[0049] As shown in Table 1, corneal epithelial cells were cultured in a medium containing 0.5% potassium sorbate. The MTT results showed that 0.5% potassium sorbate caused a significant decrease in cell survival rate compared with CK (0%) (P<0.01). This concentration had a relatively mild effect and could be used as a concentration for inducing a dry eye corneal epithelial model.

[0050] (2) Construction of in vitro reconstructed 3D corneal epithelial model;

[0051] Culture medium I comprises: a basic component of KSFM, and additional components of 5% fetal bovine serum, 10 μg / mL insulin, 1% penicillin-streptomycin, and 1% L-glutamine, wherein the percentages are by volume.

[0052] Culture medium II comprises: the basic component is KSFM, and the added components are: 1% bovine pituitary extract, 10 μg / mL insulin, 5 ng / mL hydrocortisone, 2 ng / mL epidermal growth factor, 3 ng / mL basic fibroblast growth factor, 5 μg / mL transferrin, and 1% penicillin-streptomycin, and the percentages are volume percentages.

[0053] Corneal epithelial cells were collected and P6 cells were used. They were seeded into Transwell chambers coated with Coating Matrix at a seeding density of 120w / well. The chambers were placed in 6-well plates for culture. The corresponding volume of culture medium I was added inside and outside the chambers, with the liquid surface at the same level. On the 5th day of culture, culture medium II was replaced to start air-liquid interface culture, and the whole culture process lasted for 12 days.

[0054] The above experimental process obtained an in vitro 3D reconstructed corneal epithelial model. The model proliferated and differentiated to form a corneal epithelial reconstructed model with a multi-layered cell structure. Hemidesmosomes with a tight junction structure similar to that in natural tissues were formed between the cells and cell chambers on the bottom surface of the model. Inflammatory factor detection, tissue viability and morphological detection were performed respectively, and the morphological and biochemical indicators of the obtained model were observed.

[0055] (3) Inducing an in vitro reconstructed 3D dry eye model under low humidity conditions by simulating human administration

[0056] The low humidity culture conditions are 40°C, 5% CO2, and RH 30-60%.

[0057] 1) Drug preparation: 0.5% potassium sorbate, the negative control group was given 30 μL PBS, and the experimental group was given 30 μL potassium sorbate;

[0058] 2) Model preparation: Add 0.9 mL of culture medium II to a 6-well plate, divide the experimental groups, and place one in vitro reconstructed 3D corneal epithelial model obtained in (2) in each well;

[0059] 3) Model drug administration: All drug administration operations were performed in a clean bench. On the 8th day of gas-liquid culture, drug administration was simulated in the human body. Drugs were drawn up with a pipette and slowly dripped onto the surface of the model. Drugs were administered to one model at 1-min intervals, and the model was gently shaken to ensure that the drugs were evenly distributed on the surface of the model.

[0060] 4) Induction time: Administer once a day for 24 hours.

[0061] 5) Successful verification of dry eye model building;

[0062] ① Detection of inflammatory factors: From the dry eye models obtained in 1) to 4), 3 were randomly selected in each group. After the administration time of the drug-administered tissues was over, the culture bases were collected and stored in EP tubes at -80°C for the detection of the inflammatory cytokine TNF-α.

[0063] ② Tissue viability detection: Take out the model and start the washing procedure. Control each culture chamber to wash for 30 seconds, wash the model 15 times with duplex buffer solution, and wipe with sterile gauze or sterile cotton swab. Transfer to a 24-well plate containing 300μMTT working solution, incubate in the dark for 3 hours, remove the MTT working solution with a pipette, and completely immerse the dry eye model in a 24-well plate containing 1.5mL of isopropanol in each well. Shake on a shaker for 2 hours to obtain an isopropanol extract, and draw 200μL of the isopropanol extract into a 96-well plate and read the absorbance at 570 nm on a microplate reader.

[0064] ③ Histological structure detection: Fix with 4% paraformaldehyde and send to a third-party testing agency to detect the histological structure of the model.

[0065] The above experimental process obtained an in vitro 3D reconstructed dry eye model, and the inflammatory factors, tissue vitality and morphology were detected, and the morphology and biochemical indicators of the obtained dry eye model were observed.

[0066] (4) Verification of the effectiveness and stability of the in vitro reconstructed 3D dry eye model

[0067] ① Randomly select 3 dry eye models from each group obtained in (3). After the administration time of the drug administration tissue is over, take out the model and start the washing procedure. Control the washing of each culture chamber for 30 seconds, wash the model 15 times with duplex buffer solution, and wipe it with sterile gauze or sterile cotton swab.

[0068] ② After washing, transfer to a 6-well plate containing 0.9 mL of culture medium II and incubate for 0 h, 24 h, and 48 h, respectively.

[0069] ③ After the incubation, the dry eye model obtained in the above experimental process after incubation for different time periods was subjected to corresponding testing, and the testing method was the same as above.

[0070] The results showed that the administration time of the in vitro reconstructed 3D dry eye model induced by simulating human administration under low humidity conditions in the method of this embodiment was 24 hours. Figure 1 HE staining of the control group showed that the epithelium was well stratified and the basal cells were arranged in a columnar shape. Compared with the control group, the basal cells of the H0 experimental group were partially missing, the organization was not tight, the number of cell layers was reduced, and vacuoles appeared. Figure 2 , Figure 3It can be seen that the results of ELISA and MTT tests showed that compared with the control CK, the expression of inflammatory cytokine TNF-α increased significantly and cell viability decreased after 24 hours of administration under low humidity conditions.

[0071] Figure 4 It can be seen that after the dry eye model was incubated for 24h and 48h respectively, there was no significant difference in the thickness of the epithelial layer between the two groups of dry eye models, the histological structure was improved but the arrangement of each layer was still irregular, the tissue structure was not tight, and vacuoles appeared; the histological structure of the group incubated for 48h was improved compared with that of the group incubated for 24h, but the structure was still relatively loose. Figure 5 , Figure 6 It can be seen that there is no significant difference in inflammatory factor TNF-α and tissue viability after incubation for different time periods. From the above morphological and biochemical indicators, it can be seen that the dry eye model has good stability and remains effective after 48 hours of construction.

[0072] Comparative Example 1

[0073] (1) Repeat steps (1) and (2) of Example 1;

[0074] (2) The difference between the induction method and Example 1 is that: Comparative Example 1 simulates normal humidity conditions and adopts a human body administration method for 24 hours to induce an in vitro reconstructed 3D dry eye model.

[0075] The normal humidity culture conditions are 37° C., 5% CO 2 , and RH 100%.

[0076] (3) The modeling process and the method for detecting whether the modeling is successful are the same as those in Example 1.

[0077] The above experimental process obtained an in vitro reconstructed 3D dry eye model induced by 0.5% potassium sorbate under normal humidity conditions. Inflammatory factors, tissue vitality and morphology were detected, and the morphology and biochemical indicators of the obtained dry eye model were observed.

[0078] The results show that: Figure 1 It can be seen that the dry eye model H1 constructed by the method of comparative example 1 has a better histological structure than H0. Compared with the H0 group, the stratum corneum structure of the H1 group model is relatively tight and there are fewer vacuoles in the living cell layer. Figure 2 , Figure 3 It can be seen that compared with the control CK, the expression of the inflammatory cytokine TNF-α increased significantly and the tissue viability decreased after 24 hours of administration under normal humidity conditions, indicating that 0.5% potassium sorbate also caused damage to the model. From the above morphological and biochemical indicators, it can be seen that the dry eye model constructed using Example 1 is superior to the dry eye model constructed in Comparative Example 1, and the difference between them seems to be related to dry and low humidity conditions. Therefore, dry and low humidity seem to reproduce the dry environmental conditions that cause eye discomfort and inflammation.

[0079] Comparative Example 2

[0080] (1) Repeat steps (1) and (2) of Example 1;

[0081] (2) Inducing an in vitro reconstructed 3D dry eye model under low humidity conditions by simulating human administration.

[0082] The low humidity culture conditions are 40°C, 5% CO2, and RH 30-60%.

[0083] (3) The difference from Example 1 is that: Comparative Example 2 adopts a human body-simulating administration method under low humidity conditions, and the administration is carried out for 48 hours to induce an in vitro reconstructed 3D dry eye model. The specific method is as follows:

[0084] 1) Drug preparation: 0.5% potassium sorbate, the negative control group was given 30 μL PBS, and the experimental group was given 30 μL 0.5% potassium sorbate;

[0085] 2) Model preparation: Add 0.9 mL of culture medium II to a 6-well plate, divide the experimental groups, and place one in vitro reconstructed 3D corneal epithelial model obtained in (2) in each well;

[0086] 3) Model drug administration: All drug administration operations were performed in a clean bench. On the 8th day of gas-liquid culture, drug administration was simulated in the human body. Drugs were drawn up with a pipette and slowly dripped onto the surface of the model. Drugs were administered to one model at 1-min intervals, and the model was gently shaken to ensure that the drugs were evenly distributed on the surface of the model.

[0087] 4) Induction time: Administer once a day for a total of 2 times for 48 hours;

[0088] (4) The model detection method is the same as that in Example 1.

[0089] The above experimental process obtained an in vitro reconstructed 3D dry eye model, and the inflammatory factors, tissue vitality and morphology were detected, and the morphology and biochemical indicators of the obtained dry eye model were observed.

[0090] The results show that: Figure 1 It can be seen that by using the method of comparative example 2, it can be seen that the number of vacuoles between the layers of the H2 model in the experimental group increased and increased, each layer was greatly damaged, the epithelial cell stratification was disordered, and the number of live cells in the live cell layer was significantly reduced. Figure 2 , Figure 3It can be seen that TNF-α expression reaches a maximum value at 48 hours, and is significantly higher than the level in untreated CK tissue, and the model tissue viability decreases. From the above morphological and biochemical indicators, it can be seen that the dry eye model constructed by Example 1 is superior to the dry eye model constructed by Comparative Example 2. Continuous administration for 48 hours seriously destroys the structure and function of the model, causing irreversible damage. This method is not suitable for inducing a dry eye model.

Claims

1. A method for constructing an in vitro reconstructed 3D dry eye model induced by potassium sorbate under low humidity conditions, characterized in that: The following steps are involved: (1) Screening of potassium sorbate induction concentration; Corneal epithelial cells were collected and cultured at 0.5 x 10 4 pcs / hole~1 x 10 4 The cells were inoculated into a 96-well plate at a density of 100 μL / well. When the cell fusion rate reached 70% to 80%, the culture medium containing 0, 0.25%, 0.3%, 0.5%, 0.6%, and 1% potassium sorbate was replaced. After 24 h, the cells were incubated in the dark for 3 h with MTT working solution. Finally, 100 to 200 μL of isopropanol was added to each well and the absorbance at 570 nm was read on a microplate reader. (2) Construction of in vitro reconstructed 3D corneal epithelial model; Corneal epithelial cells were collected and cells of passages P3 to P7 were used. They were seeded into Transwell chambers coated with CoatingMatrix at a seeding density of 100±50w / well. The chambers were placed in 6-well plates for culture. The corresponding volumes of culture medium I were added inside and outside the chambers, with the liquid levels at the same level. On the 5th day of culture, culture medium II was replaced to start air-liquid interface culture, and the whole process of culture was 10-15 days. (3) Inducing an in vitro reconstructed 3D dry eye model under low humidity conditions by simulating human administration; The low humidity culture conditions are 30-40°C, 5% CO2, RH30-60%; 1) Drug preparation: Potassium sorbate with a mass volume ratio of 0.5%, 30 μL potassium sorbate is administered; 2) Model preparation: Add 0.9 mL of culture medium II to a 6-well plate, divide the experimental groups, and place one in vitro reconstructed 3D corneal epithelial model obtained in (2) in each well; 3) Model drug administration: All drug administration operations were performed in a clean bench. On the 7th to 9th day of gas-liquid culture, drug administration was simulated in the human body. Drugs were sucked up with a pipette and slowly dripped onto the surface of the model. Drugs were administered to one model every 30 s to 2 min, and the model was shaken to ensure that the drugs were evenly distributed on the surface of the model. 4) Induction time: Administer once a day for 24 hours; 5) Successful verification of dry eye model building; ① Inflammatory factor detection: The dry eye model obtained from 1) to 4) is used for the detection of inflammatory cytokine TNF-α; ② Tissue vitality detection; ③ Histological structure detection; (4) Verification of the effectiveness and stability of the in vitro reconstructed 3D dry eye model; Medium I contains: the basic component is KSFM, and the supplementary components are: 1-10% fetal bovine serum, 1-20 μg / mL insulin, 0.5-2% penicillin-streptomycin, 0.5-2% L-glutamine, and the percentages are by volume; Culture medium II contains: the basic component is KSFM, and the added components are: 0.1-5% bovine pituitary extract, 1-20 μg / mL insulin, 1-10 ng / mL hydrocortisone, 0.5-10 ng / mL epidermal growth factor, 0.5-10 ng / mL basic fibroblast growth factor, 1-10 μg / mL transferrin, 0.5-2% penicillin-streptomycin, and the percentages are volume percentages.

2. The method for constructing an in vitro reconstructed 3D dry eye model induced by potassium sorbate under low humidity conditions according to claim 1, characterized in that: The method for tissue viability detection is as follows: take out the model and start the washing procedure, control each culture chamber to wash for 30 seconds, wash the model 15 times with Dulbecco's phosphate buffer solution, wipe with sterile gauze or sterile cotton swab; transfer to a 24-well plate containing 300μMTT working solution, incubate in the dark for 3 hours, remove the MTT working solution with a pipette, completely immerse the dry eye model in a 24-well plate containing 1.5mL of isopropanol in each well, shake on a shaker for 2 hours to obtain an isopropanol extract, and aspirate 200μL of the isopropanol extract into a 96-well plate and read the absorbance at 570 nm on a microplate reader.

3. The method for constructing an in vitro reconstructed 3D dry eye model induced by potassium sorbate under low humidity conditions according to claim 1, characterized in that: The effectiveness and stability verification of the in vitro reconstructed 3D dry eye model includes the following steps: ① After the administration time of the drug-administered tissue in the dry eye model obtained in (3) is over, take out the model and start the washing procedure. Control the washing of each culture chamber for 30 seconds, wash the model 15 times with Dulbecco's phosphate buffer solution, and wipe it with sterile gauze or sterile cotton swab; ②After washing, transfer to a 6-well plate containing 0.9 mL of culture medium II and incubate for 0 h, 24 h, and 48 h respectively; ③After the incubation, the obtained dry eye model is subjected to corresponding tests, including inflammatory factor detection, tissue vitality detection, and histological structure detection.

Images