Effect of Urolithin A on the Improvement of Circadian Rhythm Dysregulation in Intestinal Barrier Induced by Inflammation

UA for the cell experiments was purchased from Sigma-Aldrich (St. Louis, MI, USA). Dimethyl sulfoxide (DMSO) was purchased from Nacalai Tesque Inc. (Kyoto, Japan). ML385 was purchased from Selleck Chemicals (Houston, TX, USA).

Caco-2 cells (ATCC# HTB-37TM) were provided by the RIKEN BRC through the National BioResource Project of the MEXT, Japan, and HT-29 cells (ATCC # HTB-38TM) were provided by the European Collection of Authenticated Cell Cultures (ECACC; Salisbury, United Kingdom). Cells were grown in DMEM (WAKO, Tokyo, Japan) supplemented with 10% (v/v) FBS (HyClone, Ottawa, Canada) and 1% (v/v) Penicillin–Streptomycin (WAKO, Tokyo, Japan). Cells were incubated at 37 °C and 5% CO₂. In this study, the cells were passaged once a week, and medium was changed every three days. Before use in experiments, cells were passaged three times following recovery from stock.

Caco-2 cells (6.0 × 10⁴ cells/well) and co-culture systems (Caco-2 cells (4.8 × 10⁴ cells/well) and HT-29 cells (1.2 × 10⁴ cells/well)) were cultured in 24-well plates (VIOLAMO (AS ONE), Osaka, Japan) until reaching 80% confluency. The gene expression rhythm of the cells was synchronized using treatment with 100 nM DEX (Nacalai Tesque, Kyoto, Japan) for 2 h. After the 2 h DEX treatment, the culture medium was replaced with fresh DMEM with FBS and PS. After 24 h culture, cells were treated with vehicle (0.1% DMSO) or UA (40, 100 μM) for 30 h. In this study, 24 h after the end of DEX treatment was defined as Circadian Time 0 (CT 0). At CT 0, the Caco-2 cells and co-culture system were treated with TNF-α (5 ng/mL) and IL-17A (1 ng/mL) as the IBD mimic condition. After 30 min of the IBD-conditioned treatment, the medium of the caco-2 cells or co-culture system was changed to DMEM with DMSO, UA, or UA+ML385 to inhibit the Nrf2 pathway.

Total RNA extraction was essentially as in Du, Y et al. (2024), with minor modifications [18]. Total RNA was isolated using QIAzol Lysis Reagent (QIAGEN, Hilden, Germany) according to the QIAzol protocol. The total RNA was reverse transcribed using the ReverTra Ace qPCR RT Master Mix with the gDNA Remover kit (Toyobo, Osaka, Japan). Subsequently, 1.2 µL of the transcribed cDNA sample (diluted to 5 ng/μL) was combined with 9.0 µL of the qPCR reaction mixtures. These mixtures contained 300 nM gene-specific primers (Integrated DNA Technologies, Singapore) and Thunderbird SYBR qPCR mix (Toyobo, Japan). Gene expression was analyzed using the Roche LightCycler^® 96 System (Roche Diagnostics GmbH, Mannheim, Germany). Gene expression copy numbers were calculated using the standard curve method with GAPDH (cell sample) and 18srrna (mouse sample) as reference genes, and were normalized with an untreated control. All primers used for RT-qPCR analysis are listed in Table 1.

Seven-week-old female C57BL/6J mice, n = 50, were obtained from Sankyo Labo Service Corporation (Tokyo, Japan). The mice were divided into 2 groups by simple randomization, and were assigned a numerical identifier (e.g., CT0 C1). Mice were bred in specific pathogen-free (SPF) facility at 22 ± 2 °C and 60 ± 5% humidity under a 12 h light (07:00–19:00)–dark cycle (19:00–07:00). Zeitgeber time 0 (ZT0) was designated as lights-on time and ZT12 as lights-off time. AIN-93M was used to feed the mice, being a common diet. Experiments were performed in non-blinded conditions. This study was approved by the Committee for Animal Experimentation at the Tokyo Institute of Technology (approval number: D2022006, approved on 24 June 2022). Animals were treated in accordance with the committee’s guidelines.

UA (Jinan Feiteng Pharmaceutical Technology Co., Jinan, China) was diluted in a mixture of 0.5% carboxymethyl cellulose (CMC) and 0.1% TW80 to a suspension of 2 mg/mL, and was stored at 4 °C. The mice were weighed daily, and were orally administered UA emulsion using a flexible oral gavage tube (Natsume Seisakusho, Tokyo, Japan), with a dosing amount of 20 mg/kg/day (n = 25). Oral administration allows for more precise dosing of Urolithin A compared to mixing it into the diet. The vehicle treated group was orally administered an equivalent volume of the mixture of 0.5% CMC and 0.1% TW80 daily (n = 25). After one week of oral administration of UA, the drinking water of the mice was changed to a 4% Dextran Sulfate Sodium (DSS) solution for 3 days to induce inflammation. If the animal died during DSS treatment, this prevented the collection of data.

Mouse fresh feces were collected to measure fecal IgA concentration. Fresh stool was collected into a 100 μL PBS containing microcentrifuge tubes and incubated on ice. Then, the concentration of stool was changed to 20 mg/mL and homogenized at 13,500× g for 10 min. The fecal supernatant was collected and stored at −80 °C for the ELISA assay.

A fecal IgA concentration ELISA assay was performed essentially as in Du, Y et al. (2024) [18]. The concentration was calculated using an IgA Mouse Uncoated ELISA Kit (Thermo Fisher Scientific, Waltham, MA, USA), according to the manufacturer’s protocols.

After being euthanized, the mice colon and the SCN parts were collected. Tissues were placed into 1.5 mL microcentrifuge tubes with 400 μL QIAzol Lysis Reagent (QIAGEN, Hilden, Germany) in an ice box. After being homogenized using a micro homogenizer, samples were centrifuged at 12,000× g for 15 min at 4 °C. The supernatants were stored at −80 °C for mRNA extraction.

The data in this study are graphically represented as mean values with standard errors and were statistically analyzed with GraphPad Prism (version 9.30; GraphPad Software, Boston, MA, USA). We used the Kolmogorov–Smirnov test or the F-value test or Bartlett’s test to evaluate the normality of data distribution or bias changes. In addition, a Parametric analysis was performed using a one-way ANOVA and Student’s t-test. Statistical significance was set at p < 0.05, and p < 0.2 data are represented.

In order to verify the circadian rhythmicity of gene expression levels, data were also analyzed using nonlinear regression fitting with the cosinor equation using GraphPad Prism [19].

To investigate whether UA influences the clock gene expression rhythms in the intestinal epithelium, Caco-2 cells were used as an intestinal epithelial cell model and clock genes, BMAL1, and PER2 mRNA expression rhythms were examined. Circadian rhythm analysis data of each group are shown in Table 2.

qPCR data suggest that BMAL1 and PER2 mRNA showed significant circadian oscillation in Caco-2 cells (Figure 1A,E). The UA treatment increased the curve fitting of BMAL1 and PER2 expression rhythms based on the R² value (Table 2). The UA treatment reduced the BMAL1 expression level observed in Caco-2 cells (Figure 1B). The result showed that the amplitude of BMAL1 tended to be advanced after 100 μM of UA treatment compared with 40 μM of UA treatment (Figure 1C). The PER2 expression level was not influenced by UA treatment (Figure 1F). However, the amplitude of the PER2 expression rhythm tended to be advanced after 40 μM of UA treatment (Figure 1G). The acrophase analysis showed that UA treatment may advance the acrophase of BMAL1 and delay that of PER2 in Caco-2 cells, but a significant difference was not detected (Figure 1D,H, Table 2).

To investigate the impact of IBD on the circadian rhythms of the intestinal barrier, an in vitro IBD mimic condition was setup by treating Caco-2 cells with pro-inflammatory cytokines TNF-α and IL-17A [20], high levels of which are a key marker of immune dysregulation in IBD patients. The expression level of clock genes and tight junction genes in Caco-2 cells were examined to confirm whether UA treatment has beneficial effects on circadian rhythm dysregulation induced by inflammation in the intestinal barrier. Circadian rhythm analysis data of each group are shown in Table 3 and Table 4.

In our IBD model, BMAL1 mRNA expression levels showed a decrease after the TNF-α + IL-17A treatment, while PER2 expression rhythm did not show any change (Figure 2A,E). UA treatment improved the curve fitting of BMAL1 expression rhythm in TNF-α + IL-17A-treated Caco-2 cells based on the R² value (Figure 2A, Table 3). However, UA treatment did not improve the BMAL1 expression level in the IBD model (Figure 2B). UA treatment may advance the acrophase of BMAL1 and increase the amplitude of BMAL1 compared with the TNF-α + IL-17A-treated group, but did not show a significant difference (Figure 2C,D, Table 2). Additionally, UA treatment improved the curve fitting of the PER2 expression rhythm based on the R² value, and showed a significant increase in the baseline of the PER2 expression rhythm (Figure 2E,F, Table 3). A higher dose, 100 μM of UA treatment, also significantly delayed the acrophase of the PER2 expression rhythm (Figure 2H, Table 3). In Caco-2 cells and HT-29 cells’ co-culture system, UA treatment also improved mRNA expression of CLOCK, which was enhanced by TNF-α + IL-17A treatment at CT12 (Figure S1A,B). However, although UA treatment showed effects on CRY1 and RORA, these results may not provide conclusive evidence for UA’s role on peripheral clock improvement in the DSS-induced IBD model (Figure S1C–F).

Tight junction gene expression results showed that CLDN1 and OCLN expression exhibited 24 h oscillation in Caco-2 cells (Figure 3A,D). TNF-α + IL-17A treatment led to a decrease in the curve fitting of OCLN expression rhythm. However, it increased the curve fitting of CLDN1 based on the R² value (Figure 3A,D, Table 4). The TNF-α + IL-17A treatment also raised the baseline and amplitude of CLDN1 expression rhythm (Figure 3E). UA treatment demonstrated a significant downregulation of the CLDN1 expression rhythm curve fitting increase caused by pro-inflammatory cytokines, while significantly lowering the baseline and amplitude of CLDN1 (Figure 3D–F). Concurrently, UA treatment tended to upregulate the baseline and curve fitting of OCLN (Figure 3B, Table 4). UA treatment may also increase the amplitude of OCLN compared with the TNF-A + IL-17A group, but did not show significance in its difference (Figure 3C).

In the next experiment, 4% DSS water treatment freely fed for 3 days was used to establish an acute DSS-induced colitis mice model [15]. Mice were treated with 20 mg/kg/day of UA or 0.5% CMC+0.1%TW80 solution as a vehicle treatment at ZT 3 for one week through oral administration before DSS treatment. Mouse feces were collected before and under DSS treatment, and the fecal inflammatory marker IgA was used to determine the effect of UA pre-treatment on IBD. On the final day, colon samples from the mice were collected, and the expression levels of tight junction genes, together with clock genes, were evaluated. The circadian rhythm data of each group are shown in Table 5 and Table 6.

Fecal IgA results exhibited that mice treated with UA for one week showed an increase in fecal IgA concentration and an improvement in the IgA expression rhythm curve fitting based on R² value (Figure 4A, Table 5). UA treatment induced a fecal IgA concentration increase during the dark phase (Figure 4A). Under DSS treatment, the concentration of fecal IgA in the Vehicle-treated group was decreased (Figure 4C). However, the fecal IgA concentration showed an upward trend after UA pretreatment compared with the Vehicle treated group (Figure 4C–E).

Colon sample data show that UA pretreatment did not affect the circadian rhythm of Bmal1 and Per2 mRNA expression in the colon (Figure 5A,B). UA treatment may advance the acrophase of Bmal1 expression rhythm in the colon. However, it did not show a significant difference (Figure 5C, Table 6). UA pretreatment appeared to improve the curve fitting of Tjp1 expression in the colon after DSS treatment. However, it did not change the baseline and amplitude of the Tjp1 expression rhythm (Figure 5D–F, Table 6). Interestingly, in this model, UA pretreatment showed a downregulation in the curve fitting of Cldn1 and Cldn4 expression (Figure 5G,J, Table 6). UA pretreatment also downregulated the expression level of Cldn1 in the dark phase and Cldn4 in the light phase (Figure 5G,J, Table 6).

To determine whether UA pretreatment could affect the central clock, samples from the SCN and surrounding brain tissue were also collected. The circadian rhythm data of each group are shown in Table 7. The results indicate that both Bmal1 and Per2 exhibited 24 h expression oscillations, and their circadian rhythms were similar with those observed in the colon samples (Figure 6A,E). After UA pretreatment, the expression levels of Bmal1 and Per2 were upregulated (Figure 6B,F). Examination of mRNA expression of other clock-related genes like Clock, Cry1, Per3, and Rev-erbα in mouse SCN also suggested that UA treatment has significantly upregulated clock related gene expression levels (Figure S2A–D). Additionally, UA treatment appeared to advance the acrophase of clock genes, although this effect was not significant (Figure 6D,H, Table 7).

According to recent study, the protective effect of UA on the intestinal barrier may be related to the AhR-Nrf2 pathway [17]. ML385 is a novel and specific antagonist of Nrf2 and inhibits the downstream target gene expression of Nrf2. In addition, it was reported that ML385 did not affect growth of lung epithelial cells as high as 25 μM [21]. In this experiment, 25μM ML385 was used to assess the role of the Nrf2 pathway in UA circadian rhythm management. An intestinal epithelial co-culture model was established with Caco-2 cells and HT-29 cells [18] and was stimulated with a combination of cytokines, TNF-α, and IL-17A to mimic the IBD condition. Initially, the rhythmic expression rhythms of the BMAL1, PER2, and CLDN1 were examined in this study. Circadian rhythm data of each group are shown in Table 8.

The results indicate that the expression rhythm of BMAL1 was dysregulated after TNF-α + IL-17A treatment and ML385 treatment (Figure 7A). After UA treatment, a recovery in the expression rhythm of BMAL1 was detected based on R² value (Figure 7A, Table 8). The treatment of UA and ML385 decreased the expression level of BMAL1 (Figure 7B). In contrast, compared with the UA group, UA and ML385 co-treatment delayed the acrophase of the BMAL1 expression rhythm (Figure 7D, Table 8). The results for PER2 exhibited that UA treatment increased the amplitude of the PER2 expression rhythm and cause a phase delay compared with IBD group (Figure 7G,H). After co-treatment with ML385, the baseline and amplitude of UA-improved PER2 expression rhythm decreased, and the phase delay induced by UA treatment disappeared (Figure 7E–H, Table 8). Treatment with the ML385 led to an increase in the expression level of CLDN1 in the co-culture system and disrupted its expression rhythm (Figure 7D–F). In the group treated with both UA and the ML385, the expression level of CLDN1 decreased, and its expression rhythm was restored (Figure 7D–F, Table 8).