NEAT1-Mediated Intestinal Barrier Dysfunction Through miR-29b-3p Binding in Lipopolysaccharide-Damaged Caco-2 Cells

Summary

This study evaluated NEAT1 as a diagnostic marker and regulator of intestinal barrier dysfunction in diarrhea-predominant irritable bowel syndrome. NEAT1 was elevated in patient serum, bound miR-29b-3p, and its knockdown improved barrier, apoptosis, and inflammatory readouts in lipopolysaccharide-damaged Caco-2 cells.

Abstract

The pathogenesis of irritable bowel syndrome with diarrhea (IBS-D) is unclear, and available therapies remain limited. This study evaluated the diagnostic value of NEAT1 in IBS-D and explored its potential regulatory mechanism. A total of 116 patients with IBS-D and 116 healthy controls were included. Serum NEAT1 and miR-29b-3p levels were detected by reverse transcription quantitative real-time PCR (RT-qPCR), and receiver operating characteristic (ROC) curves were used to evaluate diagnostic value. For mechanistic studies, a lipopolysaccharide (LPS)-treated Caco-2 intestinal epithelial cell model was established. Cell viability was assessed by Cell Counting Kit-8 (CCK-8) assay, apoptosis by flow cytometry, barrier integrity by transepithelial electrical resistance (TEER), and mRNA levels of ZO-1, occludin, and claudin-2 by RT-qPCR. Dual-luciferase reporter (DLR) and RNA immunoprecipitation (RIP) assays were used to verify the binding relationship between NEAT1 and miR-29b-3p. NEAT1 was highly expressed, and miR-29b-3p was lowly expressed in patients with IBS-D, and their levels showed a strong negative correlation. NEAT1 knockdown increased miR-29b-3p levels, enhanced cell viability and TEER, reduced apoptosis, increased ZO-1 and occludin mRNA levels, decreased claudin-2 mRNA levels, and alleviated barrier injury-associated readouts in LPS-treated Caco-2 cells. Together, these results indicate that serum NEAT1 is a candidate biomarker for IBS-D and that the NEAT1/miR-29b-3p axis may be involved in intestinal barrier dysfunction.

Introduction

Irritable bowel syndrome with diarrhea (IBS-D) is a common functional gastrointestinal disorder marked by abdominal pain, abdominal distension, and altered bowel habits¹. At present, the pathophysiological mechanism of IBS-D has not been fully elucidated². The treatment methods are mostly symptomatic relief, with limited efficacy and a high recurrence rate³. IBS-D is generally believed to involve the interaction of multiple factors, such as brain-gut axis dysregulation, visceral hypersensitivity, low-grade intestinal inflammation, and barrier dysfunction^4,5. Although these criteria have standardized diagnosis to some extent, they are subjective and may lead to diagnostic delays or misclassification, particularly when patient symptoms overlap with those of inflammatory bowel disease or other organic diseases. In routine clinical practice, there are currently no validated objective biomarkers for IBS-D. Therefore, exploring a novel molecular target that can serve as a diagnostic and has potential therapeutic value has become an urgent need in the current IBS-D research field.

In recent years, long non-coding RNAs, as key molecules in epigenetic regulation, have demonstrated great potential as diagnostic markers and therapeutic targets across a variety of diseases^6,7. The main cause of IBS-D is impairment of intestinal barrier function⁸. The destruction of tight junction proteins leads to "intestinal leakage", which in turn activates the mucosal immune response^9,10. The activation of the immune system releases numerous pro-inflammatory factors that can cause intestinal dysfunction by activating the intestinal immune-nervous system. Studies have indicated that some lncRNAs can regulate the intestinal epithelial barrier^11,12. NEAT1 has been reported to be upregulated in the colonic mucosal tissue of patients with colitis¹³. Inhibition of LncRNA NEAT1 can alleviate dysfunction in intestinal epithelial cells (IECs) in ulcerative colitis¹⁴. However, the expression profile of NEAT1 in IBS-D and its diagnostic value have not yet been investigated. Furthermore, NEAT1 may have potential as a diagnostic biomarker¹⁵. If NEAT1 is abnormally expressed in the serum of patients with IBS-D, this could complement the current Rome IV criteria by providing a molecular diagnostic tool. Meanwhile, the miRNA expression profile has also changed significantly in patients with IBS. For example, the downregulation of miR-29b-3p expression has been associated with the pathogenesis of IBS¹⁶. However, the upstream regulatory mechanisms governing miR-29b-3p expression in IBS-D remain entirely unexplored. In particular, the role of NEAT1 as a competitive endogenous RNA (ceRNA) for miR-29b-3p in the context of IBS-D has never been investigated, although this mechanism has been well established in other diseases^17,18. Furthermore, while previous studies have primarily focused on tissue-level expression, the potential of circulating NEAT1 as a non-invasive serum biomarker for diagnosing IBS-D remains entirely unknown.

We hypothesize that lncRNA-NEAT1 may participate in regulating intestinal barrier function in IBS-D by acting as a ceRNA that sponges miR-29b-3p. Based on the above research background, this study aims to: (1) evaluate the levels of NEAT1 and miR-29b-3p in the serum of IBS-D patients and their diagnostic value; (2) validate the direct binding interaction between NEAT1 and miR-29b-3p; and (3) elucidate the functional role of the NEAT1/miR-29b-3p axis in regulating intestinal barrier integrity and inflammation using an LPS-treated Caco-2 cell model.

Protocol

Inclusion of patients

All participants provided written informed consent after fully understanding the research content. The Ethics Committee of Shanghai Pudong New Area People's Hospital has reviewed and approved this research protocol (No. 2024k007). Sample size was calculated using G*Power: the selected test was a t-test, and the statistical test was designed to detect differences between two independent samples. With an effect size of d = 0.5, α = 0.05, and power = 0.9, the calculation indicated that a minimum of 86 subjects should be enrolled in each group. To account for potential unforeseen circumstances during the trial, 116 patients were enrolled in each group. A total of 116 patients with IBS-D who visited Shanghai Pudong New Area People's Hospital between July 2024 and July 2025 were included in this study as the experimental group. The inclusion criteria were: (1) IBS-D meeting the Rome IV diagnostic criteria; (2) colonoscopy indicating no organic lesions in the mucosa. Exclusion criteria: (1) Patients with inflammatory bowel disease (IBD), colorectal cancer, etc.; (2) diabetic patients; (3) patients with a recent history of abdominal surgery; and (4) patients with severe heart, liver, or kidney insufficiency. Meanwhile, 116 healthy volunteers who matched the patients' demographic characteristics, such as age and sex, were recruited as the control group. Healthy controls were confirmed to have no history of chronic gastrointestinal diseases or corresponding clinical symptoms. Five milliliters of peripheral venous blood were drawn from all subjects. The supernatant was then transferred to an ultra-low-temperature refrigerator at -80 °C for cryopreservation prior to subsequent RNA extraction.

A 10-day questionnaire was completed by each participant to record the following clinical manifestations: (a) assessment of abdominal pain severity using a 10-point visual analog scale; (b) frequency of abdominal pain during the survey period (number of painful days); (c) bowel movement frequency (maximum number of bowel movements per day); (d) Bristol Faecal Traits Scale to assess faecal traits. The subjects' emotional state was evaluated using the Hospital Anxiety and Depression Scale (HADS).

Cell culture and modeling

Caco-2 cells were routinely cultured in Dulbecco's Modified Eagle Medium (DMEM) medium containing 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin, in an incubator at 37 °C and 5% CO₂. The Caco-2 cell line spontaneously differentiates into a polarized monolayer of cells with tight junctions and microvilli, exhibiting morphological and functional characteristics similar to those of mature intestinal epithelial cells. LPS triggers an inflammatory response and disrupts tight junctions in the intestinal epithelium, increasing intestinal permeability—a process highly similar to the pathological features of IBS-D. Furthermore, given that LPS-treated Caco-2 cells have been widely used in previous studies^19,20,21 to model IBS-related barrier dysfunction and inflammation, lipopolysaccharide (LPS) treatment of Caco-2 cells was selected in this study to establish an in vitro model of IBS-D. Once the cells had grown to approximately 80% confluence, the culture medium was replaced with complete medium containing 5 µg/ml LPS²², and the cells were stimulated for 24 h. Conventionally cultured cells without LPS were used as the control group.

Cell transfection

To investigate the roles of NEAT1 and miR-29b-3p in Caco-2 cells, Caco-2 cells were transfected. The cells were divided into six groups as follows: 1. control group (no treatment); 2. LPS group; 3. LPS + si-NC group; 4. LPS + si-NEAT1 group; 5. LPS + si-NEAT1 + inhibitor NC group; and 6. LPS + si-NEAT1 + miR-29b-3p inhibitor group. Two small interfering RNAs (si-NEAT1) targeting NEAT1 were designed and synthesized: si-NEAT1_1 (5′-GCCTTGTAGATGGAGCTTGC-3′) and si-NEAT1_2 (5′-GUGAGAAGUUGCUUAGAAAUU-3′). A non-targeting random sequence siRNA (si-NC) was used as a negative control. Meanwhile, the miR-29b-3p inhibitor and its negative control were also synthesized. miR-29b-3p inhibitor: 5’-AACACUGAUUUCAAAUGGUGCUA-3’. Negative control: 5’-CAGUACUUUUGUGUAGUACAA-3’.

Caco-2 cells were plated in 24-well plates and transfected once cultures reached 60–70% confluence, following the manufacturer’s protocol. Briefly, 1.25 µL of siRNA was diluted in 25 µL of reduced-serum medium to achieve a final siRNA concentration of 50 nM. For miRNA inhibition, 2 µL of the miR-29b-3p inhibitor or its negative control was diluted in 25 µL of reduced-serum medium to a final concentration of 100 nM. Separately, 1.5 µL of the transfection reagent was mixed with 25 µL of reduced-serum medium. The diluted siRNA (or inhibitor/negative control) was then mixed with the diluted transfection reagent to form the transfection complex, which was added dropwise to the cells. Six hours after transfection, the transfection mixture was replaced with fresh complete medium, and the cells were maintained for another 48 h before downstream functional assays. Knockdown or inhibition efficiency was confirmed by RT-qPCR before further experiments.

Real-time quantitative reverse transcription PCR (RT-qPCR)

Total RNA was isolated from serum samples and Caco-2 cells subjected to the indicated treatments. Briefly, each sample was lysed with 1 mL of RNA extraction reagent and combined with 200 µL of chloroform. After centrifugation at 12,000 × g for 15 min at 4 °C, the upper aqueous phase was transferred to a fresh tube, mixed with 500 µL of isopropanol, and kept at −20 °C for 30 min. The samples were then centrifuged at 12,000 × g for 10 min at 4 °C. The resulting RNA pellet was washed twice with 75% ethanol, air-dried, and resuspended in 20 µL of RNase-free water. RNA yield and purity were measured with a micro-spectrophotometer, and only samples with A₂₆₀/A₂₈₀ values of 1.8–2.0 were used. The RNA was reverse transcribed into cDNA following the manufacturer’s protocol. For detection of NEAT1, ZO-1, occludin, claudin-2, and GAPDH, 1 µg of total RNA was used for each 20 µL reverse-transcription reaction containing 1× RT buffer, 0.5 mM dNTPs, 0.5 µM oligo(dT) primer, 200 U reverse transcriptase, and RNase-free water. The reaction was carried out at 37 °C for 15 min, followed by enzyme inactivation at 85 °C for 5 s.

Quantitative PCR was performed with cDNA as the template using a SYBR Green-based real-time PCR system. NEAT1, miR-29b-3p, ZO-1, occludin, and claudin-2 expression levels were measured with gene-specific primers. U6 served as the reference gene for miR-29b-3p, whereas GAPDH was used to normalize NEAT1 and tight junction-related genes. Relative expression was calculated using the 2−ΔΔCt method. The primer sequences are listed in Table 1.

Cell viability assay

Log-phase Caco-2 cells were plated in 96-well plates at 1 × 10⁴ cells per well. A blank control group (containing only the culture medium), a control group (untreated cells), and an experimental group (Caco-2 cells under different treatment conditions) were set up, with 5–6 duplicate wells in each group. Cells were assigned to four time points: 0, 24, 48, and 72 h. A separate culture plate was prepared for each time point to avoid repeatedly opening the incubator during measurements. After the culture period, 10 µL of CCK-8 solution was added to each well, avoiding bubble formation. Subsequently, the culture plate was returned to a 37 °C, 5% CO₂ incubator and incubated in the dark for 2 h. The absorbance of each well was measured immediately at 450 nm using a microplate reader. Ultimately, cell viability was expressed as a percentage.

Cell apoptosis assay

Apoptosis was quantified by flow cytometry after Annexin V-FITC/PI staining. Caco-2 cells from each treatment group were harvested and rinsed twice with ice-cold PBS. Centrifugation was carried out at 300 × g for 5 min at 4 °C. The cell pellet was then resuspended in 100 µL of 1× binding buffer, followed by sequential addition of 5 µL Annexin V-FITC and 5 µL PI. After gentle mixing, the samples were incubated for 15 min at room temperature in the dark. Next, 300 µL of 1× binding buffer was added, and the suspension was mixed carefully before transfer to a 5 mL flow cytometry tube. Samples were analyzed by flow cytometry within 1 h.

Trans-epithelial electrical resistance (TEER)

Permeability was assessed by measuring the TEER values of Caco-2 cells²³. Caco-2 cells were inoculated in Transwell chambers and cultured for 21 days to form a dense monolayer. Six independent biological replicates were performed under each experimental condition, and five technical replicates were also performed under each condition. Before and after the experimental treatment, the TEER values of each group were measured by a transmembrane resistance meter using electrodes sterilized with ethanol and balanced with PBS. In brief, the long rod of the electrode was immersed in the basolateral culture medium, and the short rod in the apical culture medium, to avoid contact with the filter membrane. After stabilization, the resistance value (Ω) was read, and measurements were repeated until three consistent readings were obtained. The resistance of blank Transwell filters (without cells) was measured under the same conditions, and this value was subtracted from all experimental readings. TEER (Ω·cm²) = (sample resistance – blank resistance) × effective membrane area. Data are presented as: (experimental group/control group) × 100%.

Enzyme-linked immunosorbent assay (ELISA)

The supernatants of Caco-2 cells from different treatment groups were collected after 24 h of culture. The levels of pro-inflammatory cytokines, including tumor necrosis factor-alpha (TNF-α), interleukin-6 (IL-6), interleukin-8 (IL-8), and interleukin-1β (IL-1β), were measured using commercial ELISA reagents according to the manufacturer's instructions. The assay was performed strictly in accordance with the instructions. The samples and standards were added to the wells precoated with antibodies, then incubated and washed. Then, the biotinylated detection antibody was added. After another incubation and washing, streptavidin labeled with horseradish peroxidase was added. Finally, the substrate TMB was added for color development, and the reaction was terminated with the stop solution. The absorbance of each well was measured immediately at 450 nm using a microplate reader.

Western blot

Proteins were isolated from Caco-2 cells after the indicated treatments using RIPA buffer containing protease inhibitors. Protein levels were measured with a bicinchoninic acid (BCA) assay. For each sample, 30 µg of protein was loaded per lane, resolved on 10% SDS-PAGE gels, and transferred to PVDF membranes. The membranes were blocked in 5% non-fat milk for 1 h at room temperature and then incubated overnight at 4 °C with antibodies against ZO-1 (1:1,000), occludin (1:500), claudin-2 (1:1,000), and GAPDH (1:5,000). After washing, the membranes were incubated with HRP-conjugated secondary antibodies for 1 h at room temperature. Bands were detected by enhanced chemiluminescence (ECL), and band intensity was analyzed using ImageJ. GAPDH served as the loading control.

Dual-luciferase reporter (DLR) assay

The potential binding sites of miR-29b-3p to the NEAT1 sequence were predicted through the ENCORI database (https://rnasysu.com/encori/). Subsequently, NEAT1 fragments containing wild-type (WT) or mutant (MUT) binding sites were cloned into the pmirGLO reporter vector. The constructed recombinant plasmid was co-transfected into the cells with miR-29b-3p mimics (Sense: 5’-UAGCACCAUUUGAAAUCAGUGUU-3’; Antisense: 5’-CACUGAUUUCAAAUGGUGCUAUU-3’) or inhibitors and their respective negative controls (mimic NC, Sense: 5’-UUCUCCGAACGUGUCACGUTT-3’; Antisense: 5’-ACGUGACACGUUCGGAGAATT-3’). Forty-eight hours after transfection, the cells were collected, and the activities of firefly luciferase and renilla luciferase were determined using a dual luciferase detection reagent. Renilla luciferase activity was used as an internal reference for standardization.

RNA Immunoprecipitation (RIP) assay

For the RIP assay, Caco-2 cells were harvested, rinsed with PBS, and incubated on ice for 15 min in 200 µL of RIP lysis buffer supplemented with protease and RNase inhibitors. Cell lysates were centrifuged at 16,000 × g for 10 min at 4 °C, and 10 µL of each supernatant was retained as the input control. For immunoprecipitation, 40 µL of Protein A/G magnetic beads were rotated with 5 µg of anti-Ago2 antibody or normal IgG in 200 µL of RIP wash buffer for 30 min at room temperature. After three washes with 500 µL of RIP wash buffer, 100 µL of the bead suspension was combined with 100 µL of lysate, adjusted to a final volume of 1 mL with RIP wash buffer, and incubated overnight at 4 °C with rotation. The beads were then washed five times with 500 µL of RIP wash buffer and resuspended in 150 µL of the same buffer. Protein digestion was performed by adding 15 µL of proteinase K (20 mg/mL) and 15 µL of 10% SDS, followed by incubation at 55 °C for 30 min with shaking. The supernatant was collected, and RNA was isolated with the extraction reagent and precipitated overnight. Enrichment of NEAT1 and miR-29b-3p was then assessed by RT-qPCR.

Subcellular localization analysis

Caco-2 cells were separated into nuclear and cytoplasmic fractions using the referenced commercially available nuclear-cytoplasmic separation reagent, following the manufacturer’s instructions. Total RNA was extracted from these two components, and the distribution of NEAT1 in the nucleus and cytoplasm was detected by RT-qPCR. Calibration was performed using U6 (predominantly nuclear) and GAPDH mRNA (predominantly cytoplasmic) as indicators of separation efficiency.

Statistical analysis

For clinical serum samples, a power analysis (t-test, d = 0.5, α = 0.05, power = 0.9) indicated that a minimum of 86 subjects per group was required. To account for potential unforeseen circumstances during the trial, 116 patients were enrolled in each group to detect clinically significant differences. Data normality for continuous variables was evaluated using the Shapiro-Wilk test. Because all continuous data were normally distributed (P > 0.05), parametric analyses were applied. Categorical variables were analyzed using the chi-square test. Cell-based experiments included six independent biological replicates, with five technical replicates for each biological replicate. Data are presented as mean ± SD. Comparisons between two continuous-variable groups were performed using Student’s t-test, whereas comparisons among multiple groups were analyzed by one-way ANOVA followed by Tukey’s post hoc test. Statistical significance was defined as P < 0.05.

Results

Expression of NEAT1 and diagnostic value

Compared with the healthy control (HC) group, NEAT1 expression levels in IBS-D patients were markedly higher (P < 0.0001; Figure 1A). The area under the curve (AUC) of NEAT1, as shown by the receiver operating characteristic (ROC) curve, was 0.867 (95% CI: 0.821–0.913). Its specificity was 75.9%, while its sensitivity was 84.5% (Figure 1B). This suggests that abnormal NEAT1 expression in the serum of patients with IBS-D could serve as a novel diagnostic biomarker for the condition.

Additionally, the clinical data of patients in the healthy and IBS-D groups were compared. The results showed no significant differences between the two groups in age, sex, or other characteristics (P > 0.05). However, there were significant differences in HADS, bowel movements, and stool forms. (P < 0.0001, Table 2). Regarding abdominal pain frequency, 69 patients experienced symptoms for ≥ 5 days, accounting for 59.48% of cases. Regarding the severity of abdominal pain, 72 patients had a score ≥ 5, accounting for 62.07% (Table 2). This means that patients with IBS-D are clearly distinguishable from the control group by their symptoms. Given the clinical association between elevated NEAT1 and IBS-D symptoms, we next investigated the functional impact of NEAT1 on intestinal barrier integrity using an LPS-treated Caco-2 cell model.

Effects of low levels of NEAT1 on intestinal barrier damage and inflammatory responses in LPS-treated Caco-2 cells

To model the intestinal barrier dysfunction and low-grade inflammation seen in IBS-D, we treated Caco-2 cells with LPS. When LPS was added to the cells, NEAT1 levels were upregulated compared to the control group. After transfection with si-NEAT1, the NEAT1 level was markedly decreased compared to the LPS + si-NC group (P < 0.0001, Figure 2A). LPS treatment reduced cell viability and increased apoptosis relative to the control group. After knocking down NEAT1, cell viability increased, and the apoptosis rate decreased compared with the LPS + si-NC group (P < 0.001; Figure 2B,C). After knocking down NEAT1, the LPS-reduced TEER was significantly restored (P < 0.0001, Figure 2D). LPS decreased the mRNA expression of ZO-1 and occludin and increased claudin-2 expression relative to controls, whereas NEAT1 knockdown reversed these changes (P < 0.001, Figure 2E-G). Treatment with LPS can trigger an inflammatory response in cells. The levels of the pro-inflammatory factors TNF-α, IL-6, IL-8, and IL-1β in the cells were higher than in the control group. After knockdown of NEAT1, the contents of pro-inflammatory factors in the cells were significantly decreased compared to the si-NC group (P < 0.0001, Figure 3A-D). This suggests that low NEAT1 levels reduce LPS-induced intestinal barrier dysfunction and inflammatory responses, and that NEAT1 may contribute to the progression of IBS-D.

Direct binding of NEAT1 and effects on miR-29b-3p in Caco-2 cells

In Caco-2 cells, NEAT1 is primarily found in the cytoplasm (Figure 4A). The binding sites of NEAT1 and miR-29b-3p were predicted by the ENCORI database. Based on this, we designed the WT-NEAT1 and MUT-NEAT1 sequences (Figure 4B). The DLR experimental results indicated that miR-29b-3p mimics reduced WT-NEAT1 enzymatic activity relative to the mimic NC, whereas its inhibitors increased it relative to the inhibitor NC. Conversely, neither treatment had any influence on MUT-NEAT1 (P < 0.0001, Figure 4C). The RIP assay revealed that NEAT1 and miR-29b-3p were markedly abundant in the anti-Ago2 fraction compared to the anti-IgG group, indicating a specific interaction between the two (P < 0.0001, Figure 4D). These results indicate that NEAT1 specifically binds to miR-29b-3p. Meanwhile, the miR-29b-3p level was notably decreased in IBS-D patients compared to the HC group (P < 0.0001, Figure 4E). The AUC of its ROC curve was 0.856 (95% CI: 0.808–0.905), with a sensitivity of 89.7% and a specificity of 71.6% (Figure 4F). Pearson's correlation analysis revealed an inverse correlation between NEAT1 levels and the miR-29b-3p expression (r = -0.744, P < 0.0001, Figure 4G).

Effects of low NEAT1 on intestinal barrier damage in LPS-treated Caco-2 cells and miR-29b-3p levels

In LPS-treated Caco-2 cells, miR-29b-3p expression was notably downregulated compared to the control group. Transfection with si-NEAT1 increased miR-29b-3p levels, whereas transfection with the miR-29b-3p inhibitor decreased miR-29b-3p levels again (P < 0.0001, Figure 5A). Knockdown of NEAT1 increased cell viability and decreased apoptosis, and these effects were abolished by the miR-29b-3p inhibitor (P < 0.01, Figure 5B,C). The LPS-induced reduction in TEER was restored by NEAT1 knockdown, and this restoration was again prevented by the miR-29b-3p inhibitor (P < 0.0001, Figure 5D). LPS treatment led to decreased expression of ZO-1 and occludin mRNA and increased expression of claudin-2 mRNA compared with the control group. After NEAT1 knockdown, miR-29b-3p levels increased, ZO-1 and occludin mRNA levels rose, and claudin-2 mRNA levels decreased compared with the LPS + si-NC group. After miR-29b-3p inhibition, ZO-1 and occludin mRNA levels decreased, and claudin-2 mRNA levels increased compared with the LPS + si-NEAT1 + inhibitor NC group (P < 0.001, Figure 5E-G). NEAT1 (primarily the cytoplasmic subtype NEAT1_1) acts as a sponge for miR-29b-3p. Knockdown of NEAT1 releases miR-29b-3p, thereby indirectly upregulating ZO-1 and occludin and downregulating claudin-2, which alleviates LPS-induced barrier disruption. These results suggest that, in this model, miR-29b-3p positively regulates ZO-1 and occludin expression and negatively regulates claudin-2, and that the effects of NEAT1 knockdown are mediated, at least in part, by miR-29b-3p.

We further examined the expression levels of tight junction proteins using western blot analysis. As shown in Figure 5H, NEAT1 knockdown increased the LPS-reduced expression of ZO-1 and occludin and suppressed the expression of claudin-2 compared with the LPS + si-NC group. However, co-transfection with a miR-29b-3p inhibitor weakened the protective effect of NEAT1 knockdown compared with the LPS + si-NEAT1 + inhibitor NC group. These results support the conclusion that NEAT1 knockdown restores tight junction protein expression, at least in part by upregulating miR-29b-3p, thereby improving intestinal barrier function.

Effects of low NEAT1 levels on intestinal inflammation in LPS-treated Caco-2 cells and miR-29b-3p expression

In LPS-treated Caco-2 cells, LPS triggered an inflammatory response. The levels of the pro-inflammatory factors TNF-α, IL-6, IL-8, and IL-1β were significantly higher than in the control group. After NEAT1 knockdown, miR-29b-3p levels increased, while the levels of pro-inflammatory factors decreased significantly compared to si-NC. After miR-29b-3p inhibition, the levels of inflammatory factors increased again (P < 0.0001, Figure 6A-D). Therefore, NEAT1 knockdown suppresses the release of inflammatory cytokines, at least in part, by upregulating miR-29b-3p.

These results indicate that NEAT1 levels are abnormally elevated in the serum of patients with IBS-D and that NEAT1 has a high AUC for distinguishing patients with IBS-D from healthy controls. Mechanistically, NEAT1 directly binds to and sequesters miR-29b-3p in the cytoplasm. In cellular models and functional rescue experiments, NEAT1 knockdown increased miR-29b-3p expression, improved cell survival and TEER, reduced apoptosis, and attenuated the release of pro-inflammatory cytokines. These findings support the hypothesis that NEAT1 negatively regulates intestinal barrier integrity by inhibiting miR-29b-3p and suggest that targeting NEAT1 may provide a potential diagnostic and therapeutic strategy for IBS-D.

Data availability:

The raw data underlying the figures and tables are provided as Supplemental File 1 (raw data spreadsheet) and Supplemental File 2 (unedited western blot images).

Figure 1. Expression and diagnostic value of NEAT1 in patients with IBS-D. (A) NEAT1 expression levels in serum from healthy controls (HC, n = 116) and IBS-D patients (n = 116) detected by RT-qPCR; (B) ROC curve analysis of NEAT1 for discriminating IBS-D patients from healthy controls. Data are presented as mean ± SD (n = 116 per group). **** P < 0.0001. Abbreviations: IBS-D = irritable bowel syndrome with diarrhea; ROC = receiver operating characteristic. Please click here to view a larger version of this figure.

Figure 2. Effect of NEAT1 on LPS-induced intestinal barrier damage in Caco-2 cells. (A) Transfection of si-NEAT1 reduces LPS-induced NEAT1 levels; (B) Cell viability measured by CCK-8 assay at 0, 24, 48, and 72 h. Knockdown of NEAT1 enhances cell viability; (C) Apoptosis rate determined by flow cytometry with Annexin V-FITC/PI staining. Silencing NEAT1 reduces apoptosis rates; (D) Transepithelial electrical resistance (TEER) measured in Caco-2 monolayers. Inhibiting NEAT1 increases cellular TEER; (E-G) mRNA expression levels of (E) ZO-1, (F) occludin, and (G) claudin-2 measured by RT-qPCR. Silencing NEAT1 elevates ZO-1 and occludin mRNA expression, decreases claudin-2 mRNA levels. Data are presented as mean ± SD from six independent experiments (n = 6 per group). *** P < 0.001; **** P < 0.0001, LPS compared to control, LPS + si-NEAT1 compared to LPS + si-NC. Please click here to view a larger version of this figure.

Figure 3. Effect of NEAT1 knockdown on LPS-induced pro-inflammatory cytokine release in Caco-2 cells. (A-D) Concentrations of (A) TNF-α, (B) IL-6, (C) IL-8, and (D) IL-1β in cell culture supernatants measured by ELISA. NEAT1 knockdown reduces LPS-induced TNF-α, IL-6, IL-8, and IL-1β levels. Data are presented as mean ± SD from six independent experiments (n = 6 per group). **** P < 0.0001, LPS vs. control, LPS + si-NEAT1 vs. LPS + si-NC. Please click here to view a larger version of this figure.

Figure 4. Direct binding of NEAT1 and association with miR-29b-3p. (A) Subcellular localization of NEAT1 in Caco-2 cells determined by nuclear-cytoplasmic fractionation (U6: nuclear control, GAPDH: cytoplasmic control). NEAT1 is mainly located in the cytoplasm; (B) Predicted WT and MUT binding sequences of NEAT1 for miR-29b-3p from ENCORI database; (C) The DLR assay verified the binding relationship between NEAT1 and miR-29b-3p; (D) RIP verified the binding relationship between NEAT1 and miR-29b-3p. Data in A, C, and D are mean ± SD from six independent experiments (n = 6 per group). (E) Serum miR-29b-3p expression levels in healthy controls (HC, n = 116) and IBS-D patients (n = 116); (F) ROC curve of miR-29b-3p for discriminating IBS-D patients from healthy controls; (G) Pearson correlation analysis between NEAT1 and miR-29b-3p levels in IBS-D patients. NEAT1 was negatively correlated with the expression of miR-29b-3p (r = -0.744, P < 0.0001). Data in E are mean ± SD (n = 116 per group). **** P < 0.0001. Abbreviations: WT = wild type; MUT = mutant; DLR = dual-luciferase reporter; RIP = RNA immunoprecipitation. Please click here to view a larger version of this figure.

Figure 5. Effects of NEAT1 knockdown on intestinal barrier function in LPS-treated Caco-2 cells and miR-29b-3p. (A) Knockdown of NEAT1 increases miR-29b-3p levels, whereas transfection with miR-29b-3p inhibitors reduces miR-29b-3p expression. (B) Cell viability measured by CCK-8 assay. Low levels of NEAT1 enhance cell viability by increasing miR-29b-3p; (C) Apoptosis rate determined by flow cytometry. Low levels of NEAT1 reduce the apoptosis rate by increasing miR-29b-3p; (D) Low levels of NEAT1 increase the TEER of cells by raising the level of miR-29b-3p; (E-G) mRNA expression of (E) ZO-1, (F) occludin, and (G) claudin-2. Low levels of NEAT1 increase the expression of ZO-1 and occludin and reduce the expression of claudin-2 mRNA by enhancing miR-29b-3p. (H) Representative western blot images showing ZO-1, occludin, claudin-2, and GAPDH protein expression under the indicated treatment conditions. Data are presented as mean ± SD from six independent experiments (n = 6 per group). ** P < 0.01; *** P < 0.001; **** P < 0.0001, LPS vs. control, LPS + si-NEAT1 vs. LPS + si-NC, LPS + si-NEAT1 + miR-inhibitor vs. LPS + si-NEAT1 + inhibitor NC. Abbreviation: miR-inhibitor = miR-29b-3p inhibitor; TEER = transepithelial electrical resistance. Please click here to view a larger version of this figure.

Figure 6. Effects of NEAT1 knockdown on inflammation in LPS-treated Caco-2 cells and miR-29b-3p. (A-D) Concentrations of (A) TNF-α, (B) IL-6, (C) IL-8, and (D) IL-1β in cell culture supernatants measured by ELISA. Low levels of NEAT1 reduce the content of TNF-α, IL-6, IL-8, and IL-1β in the cell supernatant by increasing miR-29b-3p. Data are presented as mean ± SD from six independent experiments (n = 6 per group). **** P < 0.0001, LPS vs. control, LPS + si-NEAT1 vs. LPS + si-NC, LPS + si-NEAT1 + miR-inhibitor vs. LPS + si-NEAT1 + inhibitor NC. Abbreviation: miR-inhibitor = miR-29b-3p inhibitor. Please click here to view a larger version of this figure.

Gene	Primer sequence
NEAT1 forward	5’-GTGGCTGTTGGAGTCGGTAT-3’
NEAT1 reverse	5’-TAACAAACCACGGTCCATGA-3’
ZO-1 forward	5’-GCCGCTAAGAGCACAGCAA-3′
ZO-1 reverse	5’-TCCCCACTCTGAAAATGAGGA-3’
Occludin forward	5’-ATGGCAAAGTGAATGACAAGCGG-3’
Occludin reverse	5’-CTGTAACGAGGCTGCCTGAAGT-3’
Claudin-2 forward	5’-GTGACAGCAGTTGGCTTCTCCA-3’
Claudin-2 reverse	5’-GGAGATTGCACTGGATGTCACC-3’
miR-29b-3p forward	5’-GCTCTAGATCAGTTACAGAAAGACCACGA-3’
miR-29b-3p reverse	5’-GCTCTAGATAGTGTCCATGCACGGACC-3’
GAPDH forward	5’-CCAGGTGGTCTCCTCTGA-3’
GAPDH reverse	5’-GCTGTAGCCAAATCGTTGT-3’
U6 forward	5’-CTCGCTTCGGCAGCACA-3’
U6 reverse	5’-AACGCTTCACGAATTTGCGT-3’

Table 1: Primer sequences.

Indicator	Control group (N=116)	IBS-D group (N=116)	P
Age, years
< 40	63	54	0.293
≥ 40	53	62
Sex
Male	64	66	0.895
Female	52	50
Smoking
NO	58	53	0.599
YES	58	63
Drinking
NO	63	58	0.599
YES	53	58
HADS, score
< 8	96	26	< 0.0001
≥ 8	20	90
Bowel movements, number/day
< 4	103	51	< 0.0001
≥ 4	13	65
Stool form, Bristol score
< 5	87	25	< 0.0001
≥ 5	29	91
Abdominal pain frequency, number of days
< 5	/	47	/
≥ 5	/	69
Severity of abdominal pain, score
< 5	/	44	/
≥ 5	/	72

Table 2: Comparison of demographic and clinical characteristics between healthy controls and IBS-D patients. Values are reported as participant counts for each group (n = 116/group). Between-group differences were evaluated using the chi-square test. Abbreviation: HADS = Hospital Anxiety and Depression Scale; IBS-D = irritable bowel syndrome with diarrhea.

Supplemental File 1: Raw data underlying the figures and tables. This file contains the raw numerical data used to generate the quantitative results shown in Figure 1, Figure 2, Figure 3, Figure 4, Figure 5, and Figure 6, and the demographic/clinical data summarized in Table 2.Please click here to download this file.

Supplemental File 2: Unedited western blot images. This file contains the unedited western blot images for ZO-1, occludin, claudin-2, and GAPDH corresponding to the representative western blot panel shown in Figure 5H.Please click here to download this file.

Discussion

IBS-D is a common functional gastrointestinal disorder. Its pathological mechanism has not been fully elucidated. In recent years, lncRNA, as key molecules in epigenetic regulation, have shown great potential in the field of disease diagnosis and treatment. LncRNA NEAT1 has been studied in various diseases, such as colorectal cancer²⁴, glioma²⁵, and myocardial infarction²⁶. Multiple studies have shown that NEAT1 is involved in regulating inflammatory responses¹³, cell proliferation²⁷, and intestinal barrier integrity²⁸. However, the specific expression of NEAT1 in IBS-D, its clinical value, and the regulatory mechanisms underlying IBS-D have not yet been thoroughly explored.

NEAT1 is upregulated in various inflammatory conditions, such as pneumonia²⁹, osteoarthritis³⁰, rhinitis³¹, and sepsis³². Our results align with those reported in these studies: NEAT1 expression is significantly upregulated in patients with IBS-D. This suggests that it may be associated with the onset of IBS-D. Currently, the diagnosis of IBS-D relies primarily on the Rome IV symptom criteria, and no objective laboratory biomarkers are available. Although these criteria have proven effective in research and many clinical settings, they still have limitations, including subjectivity, symptom overlap with other diseases, and diagnostic delays. Further ROC curve analysis indicates that NEAT1 has relatively good diagnostic value in patients with IBS-D. Serum NEAT1 expression levels effectively distinguish IBS-D patients from healthy controls, with an AUC of 0.867 and high sensitivity and specificity. This suggests that NEAT1 holds promise as a tool to supplement the Rome IV criteria for diagnosis or as a novel diagnostic marker for IBS-D. For example, patients presenting with chronic abdominal pain and changes in bowel habits who have elevated NEAT1 levels may be evaluated with greater confidence when biomarker results are considered alongside the Rome IV criteria. This facilitates the early initiation of symptom-targeted treatment and avoids unnecessary invasive examinations. Conversely, for patients with normal biomarker results but who meet the Rome IV criteria, closer monitoring is warranted to rule out other organic diseases. Therefore, integrating these biomarkers into clinical practice is expected to improve diagnostic accuracy and efficiency, although prospective validation in clinical settings is still required.

The clinical symptoms experienced by patients with IBS-D may be related to impaired intestinal barrier function³³. This function is primarily mediated by the tight junction (TJ) structure between intestinal epithelial cells, with ZO-1 and occludin proteins playing important roles^34,35. ZO-1 is involved in maintaining the mechanical barrier and permeability of the mucosal epithelium³⁶. Occludin is primarily responsible for forming tight junctions between cells, which are crucial for maintaining their barrier function³⁷. The tight junction protein claudin-2 plays an important role in forming cell bypass pore channels, among other functions³⁸. In patients with IBS-D, ZO-1 and occludin are typically reduced, claudin-2 is increased, the intestinal barrier is damaged, and permeability is enhanced. This study found that NEAT1 knockdown increased the TEER of the in vitro cell model, supporting a damaging effect of NEAT1 on the intestinal barrier in this model. At the molecular level, inhibiting NEAT1 expression led to a significant rebound in the mRNA expression levels of the key tight junction proteins ZO-1 and occludin, while claudin-2 expression decreased. This suggests that NEAT1 may negatively regulate the expression of tight junction proteins, disrupting the physical barrier of the intestinal epithelium and increasing its permeability. Ultimately, this can lead to the phenomenon of 'leaky gut'. Additionally, the low-grade inflammation widespread in IBS-D is an important factor that exacerbates barrier damage³⁹. Numerous studies have confirmed that lncRNAs can act as key regulatory nodes in inflammatory signaling pathways⁴⁰. Our research is consistent with this. Knockdown of NEAT1 significantly reduces the release of the pro-inflammatory cytokines TNF-α, IL-6, IL-8, and IL-1β induced by LPS. This is highly similar to the role of NEAT1 as a pro-inflammatory lncRNA in other inflammatory models, indicating that in IBS-D, there may be a vicious cycle of inflammation-barrier damage-inflammation, in which NEAT1 plays an important role.

MicroRNAs (miRNAs) are another important class of factors involved in disease regulation. miR-29b-3p is lowly expressed in various LPS-induced cellular inflammatory responses^41,42. Consistent with most studies, we found that miR-29b-3p levels were notably downregulated in the serum of IBS-D patients and showed a strong negative correlation with NEAT1 upregulation. We further showed that NEAT1 directly binds to and sponges miR-29b-3p in the cytoplasm in this model. The functional rescue experiment further showed that NEAT1 affects the viability and apoptosis of Caco-2 cells, as well as the levels of ZO-1, occludin, and claudin-2 mRNA, at least partly through miR-29b-3p, and ultimately influences intestinal barrier integrity. Notably, our data indicate that, in the context of IBS-associated intestinal barrier dysfunction, miR-29b-3p levels are positively correlated with ZO-1 and occludin levels. This positive regulation may be indirect; for example, miR-29b-3p may suppress the transcription of tight junction proteins by targeting their transcriptional repressors, thereby lifting repression and promoting the expression of ZO-1 and occludin. This hypothesis is also supported by previous studies, which have reported that miR-29b-3p plays a protective role in inflammatory conditions by targeting negative regulators of barrier function⁴².

However, the target genes through which miR-29b-3p regulates ZO-1 and occludin mRNA expression require further investigation to complete the ceRNA network, which is a limitation of our current study. Additionally, our study has other limitations: First, this was a single-center study with a relatively small sample size (only 116 IBS-D patients were included). The clinical diagnostic value of NEAT1 still needs further validation through large-scale, multicenter prospective cohort studies. Second, this study is mainly based on clinical association analysis and in vitro cell models. In the future, animal models will be needed to verify the role of this axis in vivo. Third, the specific downstream target gene network of miR-29b-3p involved in the pathogenesis of IBS-D has not yet been fully elucidated. Future studies should focus on: (1) validating the diagnostic value of NEAT1 and miR-29b-3p in IBS-C and IBS-M patients; (2) testing the therapeutic efficacy of NEAT1-targeting interventions in animal models; and (3) elucidating the precise molecular mechanism by which miR-29b-3p regulates tight junction proteins. Such efforts will facilitate the clinical translation of our findings.

In conclusion, NEAT1 is highly expressed in patients with IBS-D and has good diagnostic value in this cohort. Low NEAT1 levels can promote cell viability, reduce apoptosis, alleviate intestinal barrier damage, and suppress inflammatory responses by increasing miR-29b-3p levels in LPS-treated Caco-2 cells. The NEAT1/miR-29b-3p axis provides a new perspective on the complex pathophysiological mechanisms of IBS-D. NEAT1 may serve as a potential diagnostic biomarker for IBS-D, and the NEAT1/miR-29b-3p axis may represent a potential therapeutic target for the condition.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
24-well plates	Corning	CLS3524	For seeding cells
96-well plates	Corning	CLS3599	For seeding cells in CCK-8 assays
Annexin V-FITC/PI apoptosis kit	Invitrogen	V13242	For detecting apoptosis rates
Anti-AGO2 Monoclonal antibody	Proteintech	67934-1-Ig	For immunoprecipitation of Ago2-bound RNA complexes
Anti-Claudin-2 antibody	Invitrogen	32-5600	Primary antibody, dilution 1:1000
Anti-GAPDH antibody	Cell Signaling Technology	5174	Primary antibody, dilution 1:5000; loading control
Anti-Human IgG antibody	Proteintech	10284-1-AP	Used as negative control for RIP assay
Anti-Occludin antibody	Abcam	ab31721	Primary antibody, dilution 1:500
Anti-ZO-1 antibody	Proteintech	21773-1-AP	Primary antibody, dilution 1:1000
Caco-2 cells	SUNNCELL	SNL-068	For constructing intestinal epithelial barrier models
CCK-8 kit	Solarbio	CA1210	Cell Proliferation Assay Kit
Chloroform	Sigma-Aldrich	C2432	For phase separation in RNA extraction
DMEM medium	Gibco	11965092	For cell cultivation
Dual-Luciferase Reporter Assay Kit	ThermoFisher	16186	For detecting luciferase activity
ELISA kit (for IL-1β)	R&D Systems	DY201	For detecting target proteins in cell culture supernatants
ELISA kit (for IL-6)	R&D Systems	D6050B	For detecting target proteins in cell culture supernatants
ELISA kit (for IL-8)	R&D Systems	NBP3-28022	For detecting target proteins in cell culture supernatants
ELISA kit (for TNF-α)	R&D Systems	DTA00D	For detecting target proteins in cell culture supernatants
Enhanced chemiluminescence (ECL) substrate	Thermo Fisher	32106	For visualizing protein bands
Ethanol (75%)	Sigma-Aldrich	459844	For washing RNA pellet
FBS (Fetal Bovine Serum)	Gibco	10099141	For cell cultivation
Flow cytometer	Beckman Coulter	B53002	For detecting apoptosis rates
HRP-conjugated secondary antibodies	Cell Signaling Technology	7074	Catalyze the generation of a chemiluminescent signal from ECL substrates
Isopropanol	Sigma-Aldrich	1133502500	For RNA precipitation
Lipofectamine 3000	ThermoFisher	L3000015	Transfection reagent
Lipopolysaccharide (LPS)	Sigma-Aldrich	L2880	For inducing cell model construction
Microplate reader	BioTek	Epoch 2	For detecting the absorbance of samples in 96-well plates
Nanodrop micro-spectrophotometer	Thermo Scientific	ND-2000C	For determining RNA concentration and purity
Nonfat dried milk	Coolaber	CN7861	For blocking membranes (5% solution)
PARIS Kit	Thermo Fisher	AM1921	For nuclear-cytoplasmic fractionation
Penicillin–Streptomycin	Gibco	15140122	For cell growth resistance to contamination
Pierce BCA Protein Assay Kit	Thermo Fisher	23225	For determining protein concentrations
pmirGLO Dual-Luciferase vector	Promega	E1330	Dual luciferase reporter gene vector
PrimeScript RT Reagent Kit	Takara	RR047A	For reverse transcription of RNA into cDNA
PVDF membrane	Millipore	IPVH00010	For protein transfer after SDS-PAGE
Real-time fluorescence quantitative PCR instrument	Applied Biosystems	4365464	For detecting target genes
RIPA lysis buffer (with protease inhibitors)	Thermo Fisher	89901	For total protein extraction from cells
RNA-Binding Protein Immunoprecipitation Kit	Sigma-Aldrich	17-701	RNA-Binding Protein Immunoprecipitation (RIP) Assay Using Protein A/G Magnetic Beads
SDS-PAGE reagents (10% gel)	TargetMol	C0189	For separating proteins by molecular weight
SYBR Green master mix	Thermo Fisher	4309155	For use in PCR testing
TEER meter (Volt-Ohm Meter)	Millipore	MERS00002	Transmembrane Resistance Meter, Measuring Single-Layer Integrity
Transwell chambers	Corning Costar	CLS3413	For culturing cell monolayers and measuring TEER
TRIzol reagent	Invitrogen	15596026CN	For total RNA extraction