Liver kidney organoid serial co culture for drug testing

Repeated dose multi-drug testing using a microfluidic chip-based coculture of human liver and kidney proximal tubules equivalents

Translation and organization: Beijing Best Technology Co., Ltd

The multi organ microfluidic serial chip can simulate the microenvironment of tissue organoid culture, achieve serial culture between organoids, and reduce species differences. Therefore, this technology has become a promising tool for preclinical drug screening. The China National Institute for Food and Drug Control, the Institute of Safety Evaluation (National Center for Drug Safety Evaluation and Monitoring), the School of Basic Medicine of Capital Medical University, the Beijing Institute for Drug Control and Research, and the Berlin Institute of Technology in Germany have collaborated with scientists from TissUse GmbH to establish an organoid model based on microfluidic chips, which can connect liver and kidney organoids and cultivate them together for 16 days. At the same time, repeated systemic administration of cyclosporine A (CsA) alone or in combination with rifampicin was conducted for 14 days. Comparing the toxicity characteristics of two different doses of CsA on different target organs, compared with continuous 14 day treatment with CsA, the combination of rifampicin starting from day 6 will reduce CsA concentration and alleviate toxicity. The serial co culture of liver and kidney organoids on organoid chips demonstrates its potential as an effective translational tool for repeated dose multi drug toxicity screening in preclinical drug development.

Drug induced nephrotoxicity and hepatotoxicity are associated with severe incidence rate and mortality. The kidneys play a crucial role in maintaining electrolyte balance and processing filtration, secretion, and reabsorption. The renal proximal tubules actively transport (reabsorb or secrete) exogenous substances, thanks to the strong polarization of proximal tubular epithelial cells expressing functional transporters. The liver controls the metabolism of exogenous substances, while the kidneys are responsible for expelling them from the body. The plasma concentration and bioavailability of exogenous substances can be altered through liver toxicity/detoxification, leading to changes in renal toxicity. Therefore. The renal proximal tubules and liver are the main targets of toxic compounds. Currently, preclinical models used to demonstrate drug nephrotoxicity and hepatotoxicity include in vitro and in vivo models. However, due to the long cultivation time and ethical limitations of animal models, traditional cell culture methods under static conditions cannot simulate microenvironments similar to those in vivo.

In order to support the ToxCast/Tox21 program, it is necessary to explore more high-throughput toxicity assessment and prediction methods. The development of human organ chip technology has overcome the shortcomings of current preclinical models and supports the 3R principle of replacing, reducing, and optimizing animal use. The current organoid serial chip integrates microfluidic pumps and organoid culture components, making the ratio of liquid to tissue similar to that in the human body. The liver kidney tandem chip ensures the distribution and exchange of nutrients, compounds, and organoid metabolites in the system, thereby guaranteeing communication between the two types of organs.

Figure 1: Microfluidic liver and kidney two organ serial chip. (A) An enlarged view of the device, including a PDMS chip that accommodates two microchannels. (B) The bold red arrow indicates the direction of liquid flow within the microchannel, as well as the experimental setup for co culturing liver and kidney organoids in MOC. (C) Schematic diagram of 16 day co culture, including a 2-day adaptation period and 14 days of repeated administration, daily exchange of culture medium, collection of supernatant on days 1, 7, and 14, and collection of cell samples on days 0 and 14.

Balanced homeostasis ensures that different stages of liver and/or kidney damage can be observed. In addition, human renal proximal tubular cells can form a complete barrier and express functional enzymes and transporters that are crucial for drug-induced kidney injury and drug interactions. The co cultured liver organoids expressed liver specific markers, human enzymes, and transporters at levels comparable to or even higher than 2D cell culture, and maintained for at least 28 days.

Cyclosporine A (CsA) is an immunosuppressant with nephrotoxic and hepatotoxic effects, used as a model drug to stimulate chip based liver kidney co culture systems. To further evaluate drug metabolism and toxicity on the dual organ chip (2-OC), we used a combination administration of CsA and rifampicin (RFP, a liver enzyme and transporter inducer). The research results on the pharmacology, pharmacokinetics, and toxicity of CsA indicate that there are significant differences between different species, especially in terms of metabolic characteristics. Therefore, based on the dosage used in clinical trials, two different doses of CsA (non-toxic/toxic dose) and RFP treatment dose were applied to the two connected organ serial chip.

To our knowledge, this isSHOUCIThere are studies indicating that the liver kidney organoid serial co culture model based on microfluidic chips can be used for toxicity studies of repeated single or combined administration for 14 consecutive days. The liver kidney organoid chip can effectively simulate the interaction process of drugs in the human body and their impact on toxicity. The research results indicate that the liver kidney organoid serial co culture chip has the potential to combine morphology, histopathology, molecular biology, drug metabolism, and non-invasive toxicity biomarker detection to elucidate drug interactions, metabolism, and toxicity.

result

We have designed a dual organ chip that can accommodate two identical microchannels, with each circuit connecting the renal proximal tubular barrier in series with visceral organs for subsequent co culture. (Figure 1A/B).

Characteristics, integrity, and function of the renal proximal tubular barrier:RPTEC/TET1 cells were cultured on Transwell and differentiated to form a complete single-cell layer (Figure 2A). The immunofluorescence staining results showed that RPTEC/TET1 cells clearly displayed the pan epithelial marker CK8/18 (Figure 2B, C). The formation of primary cilia was observed through staining of acetylated microtubule proteins (Figure 2D). Figures 2E and 2F respectively show the localization of MRP-2 and Na+- K+- ATPase, which are indicators of functional polarized monolayer RPTEC. The expression of ZO-1 (Figure 2G) and Claudin-10 (Figure 2H) at the cell boundary confirmed the formation of tight junctions between adjacent RPTEC/TET1 cells and indicated appropriate cell polarization and barrier integrity. The positive staining of Ki67 (Figure 2I) showed proliferating cells.

The RT qPCR analysis results of functional transporters MRP-2, P-gp, and Ki67 during a 16 day culture period in 2-OCs are shown in Figures 2J-L, serving as a control for gene expression changes.

Figure 2: Characteristics of renal proximal tubular barrier. (A) Transwell forms a monolayer of renal proximal tubular cells. (B) CK8/18 (green) (C) CK8/18 (green), cross-section (D) acetylated microtubule protein (green) (E) MRP-2 (red) (F) sodium potassium ATPase (red) (G) ZO-1 (red) (H) Claudin-10 (red) (I) Ki67 (red). The cell nucleus was stained with DAPI (blue). Scale: B, D, and E: 100 μ m; C. F, G, H, and I: 50 μ m. Perform qRT PCR analysis on the gene expression of (J) MRP-2 (K) P-gp (L) Ki67. The data is the mean ± standard error, and the experiment was repeated three times.

Characteristics and functions of the liver:The morphology and differentiation of HepaRG cells, formation of hepatic glomeruli, immunofluorescence staining, and RT-PCR analysis of liver functional enzymes and transporters are shown in the appendix. Figure 1. The results indicate that the liver spheroids in the microfluidic multi organ serial chip have the characteristics of long-term maintenance and repeatable differentiation, achieving a balance between cell proliferation, differentiation, and apoptosis, thus achieving physiological balance of the liver spheroids.

Performance of 16 day cultivation of two organ serial chip:The detection results of LDH levels in renal organs (representing the excretory system) and microchannel pathways (replacing blood circulation) showed (Figure 3A) that the production levels of LDH in the excretory system and blood circulation in the two organ chip remained stable. This indicates that during the entire 16 day co culture process, cells in different culture chambers underwent artificial but stable physiological replacement.

The glucose level measurements in both culture media during the 16 day co culture period are summarized in Figure 3B. The glucose concentration in the microchannel pathway and culture chamber is strictly maintained between 2.5 mM and 6 mM to ensure sufficient energy supply for the dual organ chip. The glucose level in the proximal lumen decreased by approximately three times compared to 17.5 mM in the initial culture medium. The glucose concentration in the microchannel circulation continues to decrease to 50% of the initial 11.1 mM, which is within the normal range of blood glucose. These findings indicate that during the 16 day liver kidney co culture period, the dual organ chip established a very stable glucose balance.

picture3. Performance of liver and kidney organ co culture system within 16 days. (A) The system tissue vitality in the 2-OC system is represented by the production of lactate dehydrogenase (LDH) through microchannels (red) and proximal lumens (yellow). (B) The glucose concentration in the channel pathway (red) and the proximal lumen (yellow) of the 2-OC is balanced. The dashed line represents the initial glucose concentration of the culture medium added daily to the liver (red) and proximal lumen (yellow). (C) Measure the metabolic activity of co culture by measuring lactate production and glucose consumption in the circulating culture medium.

We used the glucose consumption and lactate production in the microchannel pathway as indicators to evaluate the metabolic activity of the co culture system (Figure 3C). The metabolic activity of the 2-OC system shows a biphasic curve. In the absence of medication in the first two days, the metabolic activity continues to decline, but from day zero onwards, the metabolic activity enters a stable state with slight fluctuations. The slight decrease in lactate production in microchannels may indicate a conditional effect between the hepatic globus and proximal tubular barrier.

Both cell viability and metabolic curves demonstrate that the 2-OC system can simultaneously and continuously culture liver and kidney like organs for 16 days, while maintaining their vitality and characteristic functions. Establishing a serial co culture system with microfluidic connections in a highly reproducible and stable manner is beneficial for the application of these substances in subsequent experiments.

A 14 day repeated dose drug exposure model for a chip based liver kidney organoid serial culture system:Treat a chip based liver kidney co culture model with CsA or CsA combined with RFP to observe the interactions between organs under different degrees of toxicity threat. On the first day (Day 1) and last day (Day 14) of administration, as well as 24 hours after the first day of combined administration (Day 7), samples were taken to detect the concentration of CsA (Figure 4A) or RFP (Figure 4B) after hepatic organ metabolism. As shown in Figure 4C, 24 hours after administration on Day 1, Day 7, and Day 14, the exposure of CsA to hepatic organs in the microchannel pathway increased with increasing dose levels from 5uM to 20 uM. The CsA level 24 hours after continuous administration for seven days was lower than the level after the first and last administration. After the first and last combination administration, the concentration of CsA in the microchannel pathway decreased when 20 uMCsA was administered in combination with 25uM RFP systemically. After 8 days of simultaneous administration of RFP, the decrease in CsA concentration between cycles with or without RFP was slightly reduced. Meanwhile, after 8 days of daily combined administration, the concentration of RFP increased (Figure 4D), indicating the accumulation of RFP in the microchannel pathway. To evaluate the effects of toxic substances on liver and kidney organs, we examined the release of lactate dehydrogenase (LDH). LDH, as a reliable biomarker for evaluating in vitro nephrotoxicity, was measured in two different culture media. On the 1st, 7th, and 14th day, as the concentration of cyclosporine A (CsA) increased, the concentration of LDH in the microchannel pathway gradually increased (Figure 5A). After 8 consecutive days of combined administration of 25uM RFP, the level of LDH in the bloodstream decreased after systemic administration of 20uM CsA, which is consistent with the concentration of CsA in the microchannel pathway. However,SHOU CINo such decrease was observed after combined administration (Figure 5A). At the endpoint of the experiment, the expression of p53 gene in the combined treatment group increased by 1.5 times compared to the untreated control group, while the expression of Ki67 remained unchanged (Figure 5B, C). However, the release of LDH in the excretory cavity showed a sharp but not significant increase (Figure 5H).SHOU CIAfter combined administration, the IDH levels continued to increase on day 7 compared to the untreated control group (Figure 5H). However, after using RFP8 simultaneously for 8 days, a sharp decrease in LDH levels was observed (Figure 5H). In addition, compared with control cells, the expression level of p53 gene in renal organs treated with high-dose CsA for 14 days increased by 2.3 times (Figure 5I, J). In renal organoids treated with both CsA and RFP, the expression level of Ki67 gene increased by 2.7 times, while at the end of the experiment, the expression level of Ki67 in both low-dose and high-dose groups significantly decreased (Figure 5I, J). As shown in Figure 5D-G, TUNEL/Ki67/DAPT triple staining showed that CsA exposure increased cell death, while alleviation of cell death was observed in the liver corpus cavernosum treated with CsA in combination with RFP. The proliferating cells observed by Ki67 positive staining were evenly distributed in the hepatoglobus and maintained in the chip, with no significant changes observed between different treatment groups (Figure 5D-G).

Protein expression of conventional and novel non-invasive toxicity biomarkers:To explore further applications of MOC platform in early detection and diagnosis of organ specific toxicity, we tested conventional and novel biomarkers approved for clinical practice to determine drug-induced liver and kidney injury in this experiment. The observation results of changes in biomarker levels are listed in Supplementary Table 3.

Figure 4: Determination of CsA and RFP concentrations and metabolism of liver kidney organoids in 2-OC after daily or combined administration. (A) Mass spectrometry of CsA in circulating medium. The arrow represents the characteristic peak of CsA. (B) Mass spectrometry of RFP in circulating culture medium. The arrow represents the characteristic peak of RFP. (C) Daily repeated administration, concentration of CsA in the circulating medium after combined administration, and metabolic status of liver kidney organs in 2-OC on days 1, 7, and 14. (D) The concentration of RFP in the circulating medium after repeated daily combined administration, as well as the metabolic status of liver kidney organoids in 2-OC on days 7 and 14. The experiment was repeated three times in the control group and low-dose group, and four times in the high-dose group and combination therapy group. *(#) or * * (# #) respectively represent significant differences with p-values less than 0.05 or 0.01. *Or # respectively represent the differences compared with the corresponding control group or combination therapy group and high-dose group.

Biomarkers in the microchannel pathway (Figure 6) showed a dose-dependent increase in liver enzyme AST, with AST levels in the combination therapy group at the endpoint rather than RFPSHOU CI showed a decrease after medication. ALP levels increased after CsA treatment, with significant changes observed only on the 14th day, while simultaneous use of RFP decreased ALP levels upon administration of 20uM CsA. A significant increase in TBiI can only be observed in the post RFP stage. The decrease of glucose in the microchannel pathway is dose-dependent, and these molecules reached statistically significant levels in both the high-dose group on day 7 and the treatment group on day 14. After simultaneous treatment with 20 uM CsA and RFP, the glucose concentration in the surrogate blood circuit increased during the combined administration phase. The renal toxicity biomarkers CRE and UREA in the substitute blood increased on day 1, but no drug-related changes were observed on days 7 and 14. No changes related to CsA were found in the levels of AB, GGT, and Lac.

Biomarkers in the excretory system: In the excretory medium (see Figure 7), GGT (gamma glutamyltransferase) levels showed a dose-dependent increase on the first day. However, after seven consecutive sessions of CsA treatment, the GGT levels in the high-dose group significantly decreased, and a similar effect was observed on day 14. Simultaneous treatment with RFP did not result in a decrease in GGT levels on days 7 and 14. The same applies to the proximal tubular barrier of all treatment groups.

Figure 5. Toxicity profile of CsA14 used alone or simultaneously daily in a liver kidney organ cascade chip. (A) LDH activity in substitute blood culture medium on days 1, 7, and 14. QRT PCR analysis of the expression of (B) p53 (C) Ki6 gene in liver spheroids at the endpoint of the experiment. At the end of the experiment, TUNEL/Ki67 (green/red) staining was performed on hepatic spheroid cells from (D) control group, (E) low-dose group, (F) high-dose group, and (G) combination therapy group. (H) LDH activity in the excretory medium on days 1, 7, and 14. QRT PCR analysis of (I) p53 and (J) Ki67 gene expression in the proximal tubular barrier at the end of the experiment. The data is the mean ± standard deviation. The experiment was repeated three times in the control group and low-dose group, and four times in the high-dose group and combination therapy group. *(#) or * * (# #) respectively represent significant differences with p-values less than 0.05 or 0.01. *Or # respectively represent the differences compared with the corresponding control group or combination therapy group and high-dose group.

As the dosage of CsA increases, NAG enzyme in urine will also be secreted. The effects of CsA and RFP on ALP in the excretory pool are similar. Under the action of CsA, glucose decreased in a dose-dependent manner, and the glucose levels in the combination therapy group were lower than those in the CsA treatment group. After CsA treatment, the concentration of Lac significantly increased, while RFP alleviated the effect of CsA on Lac secretion.

Figure 6: Protein expression of non-invasive toxic biomarkers in the microchannel pathway after daily use of CsA alone or simultaneously for 14 days in a liver kidney tandem culture chip. Protein levels (A) AST (B) ALP (C) TBiL (D) Glucose (E) CRE (F) UREA (G) AIB (H) CGT (I) lactate in the circulating medium on days 1, 7, and 14. The data is the mean ± standard deviation. The experiment was repeated three times in the control group and low-dose group, and four times in the high-dose group and combination therapy group. *(#) or * * (# #) respectively represent significant differences with p-values less than 0.05 or 0.01. *Or # respectively represent the differences compared with the corresponding control group or combination therapy group and high-dose group.
KIM-1 is a kidney specific biomarker that has been officially identified by regulatory agencies and showed a dose-dependent increase throughout the study period. The CsA mediated increase in KM-1 decreased over time, and the same was true when taking RFP. Similar results were observed in the levels of CysC and ColI IV in the excretory medium. But in the post RFP stage, RFP will enhance the effect of ColI IV. After CsA treatment, the concentrations of Clu and NGAL significantly increased, while RFP significantly reduced the levels of Clu and NGAL. In addition, in the mid stage of the study, the secretion levels of these two biomarkers reached high ZUI levels, and the decrease was also the greatest after using RFP. During the experiment, the secretion level of OA was sustained, and the OA level in the high-dose group was maintained at a relatively high level. The reduction of OA in the combination therapy group showed a significant delay. Similarly, after treatment with 20uM CsA, IP-10 increased slightly, while RFP showed a slight decrease in IP-10 on the 14th day. The dose-dependent increase in GST-a and albumin levels induced by CsA was only observed on day 1. CsA enhanced FABP-1, while the decrease produced by simultaneous use of RFP was not observed until the later stage (day 14), while REN and a1 MG showed similar drug-induced changes on days 7 and 14. On the first day, CsA can cause an increase in OPN, but it is dose-dependent, while on the seventh day, the increase in OPN is dose-dependent. Although the effect of CsA on OPN disappeared on day 14, the significant decrease in OPN induced by RFP was maintained in the post RFP stage. TIMP-1 showed a dose-dependent significant increase on day 1. However, the effect of CsA on TIMP-1 was reversed on days 7 and 14. In the post RFP stage, CsA reduces the increase in TIMP-1 levels to varying degrees. After treatment with CsA or RFP, there is a dose-dependent decrease in EGF. No drug-related changes were observed in CAIB and TFF-3 (data not shown).

Figure 7: Protein expression of non-invasive toxic biomarkers in the excretory cavity after daily use of CsA alone or simultaneously for 14 days in a liver kidney tandem culture chip. (A) CGT (B) ALP (C) NAG (D) glucose (E) lactate (F) KIM-1 (G) cystatin C protein levels (H) collagen IV (I) lectin (J) NGAL (K) osteocalcin (L) IP-10 (M) GST-a (N) albumin (O) FABP-1 (P) renin (Q) α 1-microglobulin (R) osteocalcin (S) TIMP-1 (T) excretes epidermal growth factor 14 from the culture medium on days 1, 7, and 8. The data is the mean ± SEM, and the experiment was repeated three times in the control group and low-dose group, while four times in the high-dose group and combination therapy group. *(#) or * * (# #) respectively represent significant differences with p-values less than 0.05 or 0.01. *Or # respectively represent the differences compared with the corresponding control group or combination therapy group and high-dose group.

Changes in gene expression of metabolic enzymes and transporters: In this study, we further investigated the functional enzymes and transporters expressed in two organoids and their roles in detoxification or detoxification processes (Figure 8). At the end of the experiment, the simultaneous use of RFP significantly induced the expression of MRP-2 in liver and kidney organoids. After simultaneous administration with RFP, the expression of Pgp gene in hepatic spheroids slightly increased. In contrast, the expression of P-gp gene in RPTEC/TET1 cells can be significantly induced by continuous use of cyclosporine A (CsA), while induction of P-gp gene cannot be observed in the combined treatment group.

Figure 8. Gene expression of metabolic enzymes and transporters in the hepatic glomerulus and proximal tubular barrier after daily use of RFP alone or simultaneously in a liver kidney co culture chip for 14 days. At the end of the experiment, the folding changes in mRNA expression levels of liver (A) MRP-2 (C) Pgp (E) CYP3A4 (F) BSEP and kidney (B) MRP-2 (D) P-gp compared to the control group were observed. The data is the mean ± standard deviation. The experiment was repeated three times in the control group and low-dose group, and four times in the high-dose group and combination therapy group. *(#) or * * (# #) respectively represent significant differences with p-values less than 0.05 or 0.01. *Or # respectively represent the differences compared with the corresponding control group or combination therapy group and high-dose group.

In the CsA treatment group, a statistically significant and dose-dependent decrease in the expression of CYP3A4 and BSEP was observed. The inhibitory effect of CsA on BSEP was slightly weakened by simultaneous administration of RHP, while there was no significant difference in CYP3A4 between the CsA group and the simultaneous treatment group.

Discussion

The latest research shows that organoid chip technology can reconstruct three-dimensional (3D) microenvironments, simulate the structure, mechanics, physiology, and metabolic characteristics of human organs and tissues. For example, liver organoid models established using 3D technology exhibit enhanced liver function, particularly when non parenchymal cells are integrated into the organoid. We have developed a stable and repetitive differentiation protocol for liver organoids, which includes differentiated HepaRG (parenchymal) and HHSteC (non parenchymal) cells, and is capable of reproducing the liver lobular structure, which is the functional unit of the liver. The HepaRG cell line was used as an in vitro model to predict the induction of human cytochrome P450 (P450) enzymes by drugs. In addition, simulating fluid shear stress and interstitial pressure around liver like organs can help maintain a stable microenvironment based on protein and oxygen gradients in the chip for a long time. Moreover, the expression of related proteins including ZO-1 and Claudin-10 indicates that the tight junction assembly and cell polarity of the proximal tubular cell barrier were maintained throughout the co culture process.

For a long time, the measurement of creatinine (CRE) and urea (UREA) in serum has been used to diagnose renal toxicity caused by CsA; However, as observed in this study, these biomarkers do not show secretion over a long period of time, making them insensitive and insufficient. The human RPTEC/TET1 cell line has been shown to polarize and form tight monolayers, expressing specific tissue enzymes and transporters for a prolonged period of time, and has therefore been established as an in vitro model for drug safety assessment. The existing chip based liver proximal tubule co culture model replicates long-term physiological and molecular communication between two organs and ventricles. The co cultured proximal tubular barrier equivalents in the chip expressed several biomarkers, such as KIM-1, CLU, OA, and NGAL, for long-term toxicity assessment and simulation of in vivo response to renal toxicity exposure. The FDA and EMEA have approved eight non-invasive biomarkers for kidney injury that have been fully described in vivo and can be detected in animal or human urine or blood. Among these eight qualified biomarkers, KIM-1、 Albumin, Clu, and TFF-3 are considered biomarkers associated with acute tubular degeneration.

Consistent with previous studies, other biomarkers that respond to acute CsA induced nephrotoxicity, including GST - α, NGAL, OPN, GGT, NAG, and TIMP-1 in excretory mediators, as well as CRE and UREA in surrogate blood, can be classified as biomarkers of renal tubular dysfunction. Research has shown that OA is upregulated in association with acute renal tubular epithelial injury, including CsA induced injury and in vivo tubulointerstitial fibrosis. The dose-dependent increase in OA supports the conclusion that OA is a sensitive biomarker for early proximal tubular injury. The urinary α 1-MG mainly produced in the liver may increase in both tubular and glomerular injuries. Liver specific FABP (i.e. FABP-1) is expressed in the proximal tubules. These molecules are biomarkers for acute and especially chronic renal tubular injury, therefore, these findings can partially explain the elevated levels of FABP-1 and α 1-MG observed only in the late stages of treatment. Throughout the entire study period, high-dose CsA upregulated Col IV. In addition, Col IV is expressed in the renal tubular basement membrane; Therefore, the upregulation of Col IV in the co administered group indicates the regeneration of renal epithelial cells induced by Ki67 gene expression on day 14. CALB exists in the small intestine and is inhibited in the distal renal tubules after long-term use of CsA. Therefore, the response of CALB as a biomarker for renal tubular injury in vitro and in vivo to single and repeated treatments with CsA is drug independent. This isThe SHOU CI report evaluates a range of qualified and potential non-invasive renal toxicity biomarkers on the microfluidic MOC platform, demonstrating the ability and advantages of MOC for high-throughput screening in early drug development.
The intestinal barrier and glomerular equivalents will be integrated into novel MOC systems, such as four organ chips, to improve and optimize our research. On the one hand, it is necessary to consider the low bioavailability of CsA. The low bioavailability of CsA is mainly due to liver metabolism, partly due to intestinal absorption. The decrease in bioavailability is likely to be explained by the induction of intestinal CYP450 enzymes, which is significantly greater than the induction of liver metabolism. The effect of CsA on gut liver circulation in vivo will be simulated on the MOC platform.
Conclusion
The microfluidic multi organ chip platform evaluated in this study enables the non-contact co cultivation of reproducible and well-defined hepatic spheroids and renal proximal tubule barriers within 16 days through connected culture medium circuits, with each structure being 100000 times smaller than its corresponding organ in the human body. In addition, the fluid to tissue ratio designed in microfluidic devices and 2-OC models supports sufficient inter organ communication. Finally, the co cultured tissue equivalents expressed tissue-specific markers, enzymes, and transporters, and can be challenged by CsA with hepatotoxicity and nephrotoxicity, which can be alleviated by the hepatic drug metabolism inducer RFP. At present, platforms based on microfluidic chips are suitable for standardized microstructure sizes and tissue culture formats, and can be widely used for drug screening, which is increasingly accepted by regulatory agencies.

Materials and Methods

Design and manufacture dual organ chips:The commercially available microphysiological dual organ chip 2-OC is integrated with two 96 well cell culture plugins. The manufacturing of 2-OC was carried out by TissUse GmbH according to Wagner et al.'s description. The fluid flow in the circuit is established by a pulsed micro pump on the chip based on the modification of Wu et al. The micro pump on the chip is driven by compressed air from an external control unit provided by TissUse GmbH.

Human cell source:The human HepaRG cell line and the human proximal tubular cell line RPTEC/TET1 (CRL-4031) were both purchased from ATCC (ATCC located in Manassas, Virginia, USA). The HHSteC cell line was purchased from the ScienCell research laboratory located in Carlsbad, California, USA.

Liver organoid culture:HepaRG cells were cultured in HepaRG medium (containing William's medium E, supplemented with 10% fetal bovine serum (FBS, Life Technologies; Sydney, New South Wales, Australia), 2 mM GlutaMAX, 10000 U/ml penicillin, 10000 μ g/ml streptomycin, 5 µ g/ml human insulin (Sigma Aldrich, St. Louis, USA), and 5 × 10-5 M hydrocortisone succinate (National Food and Drug Administration, Beijing, China). The cell culture medium and its supplements were purchased from Life Technologies (Carlsbad, California, USA). HepaRG cells were cultured at 37 ° C and 5% CO2, and the culture medium was changed every other day. Cultivate HepaRG cells in HepaRG medium for two weeks to induce HepaRG cell differentiation and achieve adherent growth. Then add differentiation medium containing 2% dimethyl sulfoxide (DMSO; Sigma Aldrich; St. Louis, Missouri, USA) and culture for another two weeks. HHSteC cells were expanded in star shaped cell culture medium purchased from ScienCell Research Laboratory (ScienCell, Carlsbad, California, USA). Cells are harvested for further co culture when they reach 80% adherent growth.

The formation of human hepatic spheroids follows Wagner et al.'s previous description. Simply put, 50 μ l of cell suspension containing 0.1 × 104 HHSteC cells and 2.4 × 104 HepaRG cells were placed in each 384 well spherical plate well (Corning Costar, Kennebunk, USA). Cells were cultured on a shaker in a CO2 incubator for 3 days, forming circular and compact spheres. Then transfer the sphere to a low adhesion 24 well plate (Corning Costar, Kennebunk, USA). Collect 40 spheres to form a single liver equivalent of a 2-OC system. Evaluate the morphology of differentiated HepaRG cells and the formation of hepatic spheroids under EVOS XL Core digital inverted microscope (Life Technologies, Carlsbad, California, USA).

Establishment of renal proximal tubular barrier model:RPTEC/TET1 cells were cultured in DMEM and Ham's F-12 (DMEM/F12) medium supplemented with 2 mM GlutaMAX, 10000 U/ml penicillin, 10000 μ g/ml streptomycin, 36 ng/ml hydrocortisone, 5 μ g/ml human insulin, 5 μ g/ml transferrin, 5 ng/ml sodium XIate (ITS 1 ×), and 10 ng/ml human epidermal growth factor (hEGF). The culture medium and supplements were purchased from Life Technologies, located in Carlsbad, California, USA. Hydrocortisone was purchased from Sigma Aldrich in St. Louis, USA.
Each Transwell permeable scaffold can implant 100000 RPTEC/TET1 cells and allow them to attach at night. Then, replace the culture medium in the upper chamber to remove unattached cells. Prior to the experiment, the implanted scaffolds were cultured statically in 2-OC for 4 days and under perfusion conditions for 3 days. Position the insertion membrane 100 μ m above the bottom of the culture chamber within the 2-OC to ensure free passage of the medium under the barrier of the proximal tubule. Use EVOS XL Core digital inverted microscope (Life Technologies, Carlsbad, California, USA) to monitor the cell morphology and differentiation of RPTEC/TET1.

Chip based co cultivation:Fourteen 2-OC systems were used in the experiment, co cultured in 500 microliters of circulating medium, with 100 microliters of medium added to the top of the barrier plug. The culture medium in the excretory pool is replaced daily. Half of the circulating supernatant was replaced and collected. At the end of 16 days of co culture, organ specific functional markers of liver or proximal tubular equivalents were analyzed by immunofluorescence and qRT PCR. The micro pump on the chip is set to a frequency of 0.8 Hz, resulting in a flow rate of no more than 9 microliters per minute.

Experimental reagents:CsA (CAS number: 59865-13-3) was added to repeated doses of co culture for testing. RFP (CAS number: 13292-46-1) was administered simultaneously with CsA. For the stock solution, CsA and RFP (China National Food and Drug Administration - Beijing) were dissolved in DMSO and stored in a dark place at 4 ° C. When used, they were diluted to a final concentration of 0.1% DMSO.

Drug exposure to liver kidney co culture chip:After three days of drug free co culture, CsA stock solution was diluted to corresponding concentrations of fresh culture medium, 10 µ M (low dose) and 20 µ M (high dose), and administered daily for 13 days. In the combination therapy group, 20 µ M CsA was administered daily until day 6, followed by simultaneous administration of 20 µ M CsA and 25 µ M RFP daily from day 6 to day 13. The culture medium containing 0.1% DMSO was used for control culture (see Figure 1C).

Co cultivation analysis:Supernatant sample: Cell viability is monitored by daily measurement of lactate dehydrogenase (LDH) activity in the supernatant of the liver and proximal tubular barrier compartment. Metabolic activity was detected by measuring glucose and lactate levels in the supernatant of two compartments daily using a lactate colorimetric assay kit (Biovision, Milpitas, California, USA) according to the manufacturer's instructions. On days 1, 7, and 14, liquid chromatography tandem mass spectrometry (LC-MS/MS) was used to analyze the concentrations of CsA and RFP in the surrogate blood circulation, and toxic biomarkers were detected in the supernatant of both compartments. The detection of biomarkers including LDH, glucose, aspartate aminotransferase (AST), alkaline phosphatase (ALP), gamma glutamyl transpeptidase (GGT), N-acetyl - β - D-glucosaminidase (NAG), total bilirubin (TBiL), albumin (ALB, serum), creatinine (CRE), and urea (UREA) was performed using a Hitachi 7180 automatic biochemical analyzer (Japan). The above labeled reagent kit was purchased from Wako (Japan). Osteoactive hormone (OA), glutathione S-transferase alpha (GST - α), KIM-1, fatty acid binding protein-1 (FABP-1), type IV collagen (Col IV), neutrophil gelatinase associated lipocalin (NGAL), colonin (Clu), cystatin C (Cys C), renin (REN), α 1-microglobulin (α 1-MG), osteopontin (OPN), neutrophil gelatinase associated lipocalin (NGAL), calcium binding protein (CALB), tissue inhibitor of metalloproteinase-1 (TIMP-1), interferon inducible protein-10 (IP-10), trefoil factor-3 (TFF-3), The determination of epidermal growth factor (EGF) and urinary albumin was performed using Merck Millipore's MILLIPLEX MAP human kidney injury magnetic bead plates 1 and 2. According to the manufacturer's guidance, these tests were performed on Luminex 200 in Texas, USA. The data was analyzed using MILLIPEX Analyst 5.1 software (Billerica, Massachusetts, USA).

Perform tissue culture analysis before and after administration:The hepatic spheroids and membranes containing RPTEC/TET1 cells were frozen in O.C.T. compound and stored at -80 ° C until further analysis. For staining, the slices were first fixed with acetone at -20 ° C for 10 minutes or 4% PFA for 10 minutes. Then, the cells were permeabilized with 0.05% Triton X-100, washed and blocked with 10% goat serum for 30 minutes, followed by incubation with primary antibodies against CYP3A4, BSEP, CK8/18, MRP-2, Na+- K+- ATPase, P-gp, VIM, ZO-1, CLDN 10, and Ki67 (all from Abcam, Cambridge, Massachusetts, USA) at 4 ° C overnight or at room temperature for 2 hours, or with acetylated microtubule protein (from Sigma Aldrich, St. Louis, Missouri, USA). Afterwards, the slices were washed three times and incubated with fluorescently labeled secondary antibodies for 1 hour at room temperature. Staining was performed using DAPI (Sigma Aldrich, St. Louis, Missouri, USA).
Proliferation and apoptosis were analyzed by TUNEL/Ki67 immunofluorescence staining. In short, the slices were sliced using TUNEL technology (ApopTag1 peroxidase in situ apoptosis detection kit, Merck Millipore, Stain according to the manufacturer's instructions in Darmstadt, Germany. Use DAPI for dyeing. All immunofluorescence images were obtained using Nikon Eclipse 80i digital fluorescence microscope (Tokyo, Japan) and Olympus IX 71 inverted microscope (Tokyo, Japan) in conjunction with Image Pro Plus software (version 6.0, Media Cybernetics Inc., Maryland, USA).
Total RNA was extracted from hepatic spheroids and proximal tubular cells using RNeasy @ Micro Kit (Qiagen, Dusseldorf, Germany). CDNA synthesis was performed using FastQuant RT Kit (including gDNase) (Tiangen Biotech; Beijing, China). QRT PCR amplification was performed on Line Gene 9620 model FQD-96A (Bioer) using SYBR Green assay (Tiangen Biotech; Beijing, China). The program is executed according to the manufacturer's manual for the kit and the software included in the device. Data analysis was conducted using software version 2.0.6. The changes in mRNA levels were determined using the 2- Δ Δ CT method. The primers were purchased by Sangon Biotech (Beijing, China). TBP and β - actin are reference genes for the hepatic spheroid and proximal tubular barrier. The primer sequence of the target gene is shown in Supplementary Material Table 1.

Statistical analysis:Statistical evaluation was conducted using GraphPad Prism (version 6.0, San Diego, California, USA) and SPSS (version 17.0); IBM， Conducted in Amonk, New York, USA. The P-value was calculated through one-way ANOVA and student t-test. All experiments were repeated three times (control group and low-dose group) or four times (high-dose and combination group). Unless otherwise specified, all data are presented as mean ± standard error.

Received: 18 November 2019; Accepted: 15 April 2020

Published: 1 June 2020