High palmitate induces ferroptosis in RIN-m5f cells via miR-3584-5p-mediated suppression of AQP7

Abstract

Aquaporin-7 (AQP7) is a crucial aquaporin in pancreatic islet β-cells, acting as a crucial role in sustaining cellular viability and survival. An in vitro model was employed, specifically RIN-m5f islet β-cells exposed to elevated levels of palmitic acid, to investigate the influence of high-lipid conditions on AQP7 expression and its associated downstream effects. Our findings demonstrated a marked reduction in AQP7 expression, coupled with heightened activation of oxidative stress markers and ferroptosis pathways. Notably, the downregulation of AQP7 was linked to elevated concentrations of reactive oxygen species (ROS) and lipid peroxidation, leading to impaired β-cell function. Moreover, we observed that the upregulation of miR-3584-5p under high-lipid conditions contributed to the inhibition of AQP7, thereby exacerbating oxidative stress and promoting ferroptosis. Inhibition of miR-3584-5p restored AQP7 levels, alleviated oxidative stress and improved β-cell viability. These findings reveal a novel mechanism through which AQP7 influences cellular oxidative stress and ferroptosis, emphasizing its crucial regulatory role in preserving β-cell health in high-lipid environments. Additionally, the study emphasizes the potential of targeting AQP7 and related pathways as therapeutic strategies to mitigate high-lipid-induced pancreatic β-cell injury.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-38935-4.

Introduction

Diabetes is a major global health concern, resulting in numerous deaths and the onset of chronic diseases^1–3. Type 2 diabetes mellitus (T2DM) represents the predominant form of diabetes, comprising approximately 90% to 95% of all diagnosed instances of the condition⁴. In the progression of T2DM, the functional impairment of β-cells is considered a critical determinant in the advancement of the disease. The impairment of β-cells is often closely associated with defects in insulin secretion^5,6. Excessive nutrient intake can trigger chronically elevated insulin secretion, which contributes to varying degrees to dysfunction of islet β-cells⁷. Palmitic acid (PA), a prevalent fatty acid in human physiology, can disrupt lipid metabolism and cause insulin resistance in various organs, ultimately leading to islet β-cell injury or even cell death⁸. Damage to islet β-cells is a key factor in disease progression, particularly since it is often closely linked to defects in insulin secretion^5,6. This process entails modifications in insulin release, exhaustion of insulin stores, and the differentiation of β-cells, the demise of which could significantly contribute to the advancement of the condition^9,10. Therefore, a thorough grasp of the pathways leading to pancreatic β-cell damage under hyperlipidemic conditions, coupled with the discovery of novel therapeutic targets, is urgently required.

Reactive oxygen species (ROS) are pivotal in regulating survival, proliferation, and apoptosis of β-cells^11–13. Normally, the concentration of ROS in islet β-cells remains minimal, which helps maintain cellular homeostasis and equilibrium. However, an excessive increase in ROS levels will cause pancreatic β-cells to suffer from oxidative stress, which subsequently induces dysfunction of the cell function and its death^14,15. Hyperlipidemia is strongly associated with significant increases in oxidative stress and decreases in the activity of protective enzymes, suggesting that it serves as a primary contributor to ROS generation¹⁶. In addition, palmitic acid treatment has been shown to significantly induce nitric oxide (NO) production and lipid peroxidation, whereas high-carbohydrate diet did not¹⁷. Consequently, investigating the impact of oxidative stress on pancreatic β-cell function under hyperlipidemic conditions is of critical importance. Hydrogen peroxide (H₂O₂), a major type of intracellular ROS, plays a crucial part in the cellular response to oxidative damage in pancreatic β-cells^3,18. The accumulation of H₂O₂ not only destroys the intracellular antioxidant system, but also induces apoptosis of β-cells¹⁹. Although many researches have described generation of H₂O₂ within pancreatic islet β-cells and its impact on cellular function, the precise mechanisms regulating H₂O₂ homeostasis remain poorly understood. Understanding these regulatory mechanisms is still a key focus in β-cells research.

Aquaporins (AQPs) are a group of proteins responsible for regulating the flow of water molecules through the cell membrane, regulating both the entry and exit of water in various cells. Recent studies have further shown that some AQP isoforms are not only water and small molecule transport channels, but also H₂O₂ transport channels. Although H₂O₂ and water have similar physical chemistry properties, the diffusion efficiency of H₂O₂ is relatively lower than that of water, thus more effective regulation is needed to improve the membrane permeability of H₂O₂²⁰. By transporting H₂O₂, AQP5 can facilitate cancer cell adaptation and migration, potentially important in tumor growth and metastasis. This ability to regulate oxidative stress could contribute to the aggressive behavior of certain cancers²¹. Furthermore, AQP9 plays a regulatory role during liver regeneration by transporting H₂O₂, especially after partial hepatectomy, and AQP9 can promote hepatocyte proliferation and regeneration²². These studies have shown that AQPs are not only water transport channels, but also participate in the redox balance of cells, and have the functions of regulating intracellular oxidative stress, promoting cell repair and responding to external stimuli. AQP7 is abundantly present in β-cells, where it is essential for promoting cell proliferation²³. Furthermore, AQP7 in mouse bone marrow-derived mesenchymal stem cells (BMSC) enhances cellular growth and supports the formation of adipocytes by modulating H₂O₂ transport²⁴. Previous studies have found that the application of shRNA down-regulated the expression of AQP7, can significantly down-regulate insulin secretion of islet β-cells²⁵. Although AQP7 is recognized for its involvement in the transport of water and glycerol, its role in contribution to controlling oxidative stress by regulating H₂O₂ levels in β-cells remains uninvestigated.

MicroRNAs (miRNAs) are a group of non-coding RNAs that are essential in regulating gene expression at the post-transcriptional level. They achieve this regulation primarily through the repression of translation and the degradation of messenger RNA²⁶. A growing body of research indicates that miRNAs play a part in controlling numerous physiological functions throughout various systems and cellular structures. Importantly, they are essential in critical regulatory processes, including the control of cell specialization, growth, and programmed cell death, among others²⁷. Research has shown that certain miRNAs play a key part in controlling the expression of genes vital for maintaining the stability of pancreatic islet β-cells²⁸. These miRNAs not only influence genes critical for preserving the balance of β-cells but also play key functions in the adaptive responses of β-cells to stress stimuli, the control of insulin release pathways, the preservation of β-cell population, and the development of β-cells²⁸. These studies imply that miRNAs may influence the growth and development of β-cells, or alternatively, modulate their functional activity.

Ferroptosis is a type of controlled cell death driven by iron, marked by the accumulation of lipid peroxides to toxic levels, leading to cell death. It is usually triggered by compounds that either inhibit the production of glutathione (GSH) or decrease the function of glutathione peroxidase 4 (GPX4). GPX4 is an enzyme responsible for safeguarding cells against oxidative stress by neutralizing lipid peroxides, and ferroptosis is manifested by the buildup of iron-driven ROS and the reduction of polyunsaturated fatty acids (PUFAs) at the cell membrane²⁹. Nuclear factor erythroid 2-related factor 2 (Nrf2) is the transcriptional master regulator of cellular responses against oxidative stress³⁰. The Nrf2/HO-1 signaling pathway plays a complex regulatory role in oxidative stress-related diseases³¹ and has been shown to contribute to the protective effects against ferroptosis³². Ferroptosis has also been involved in the progression of diseases associated with high-lipid diets. This is tightly connected to its involvement in the damaging processes of lipid peroxidation and the triggering of inflammation, both of which contribute to the pathogenesis of various diseases^33,34.

The aim of this study was to investigate the possible regulatory function of miR-3584-5p in ferroptosis in rat pancreatic islet β-cells following high-lipid induction, as well as examining the possible interaction between miR-3584-5p and AQP7.

Materials and methods

Cell culture

The RIN-m5f cells and 293T cells were obtained from the Cell Resource Center at the Institute of Basic Medical Sciences (IBMS). Cells were grown in RPMI-1640 (Gibco, USA) at 37 °C with 5% (v/v) CO₂ and 95% humidity. Both the palmitate-treated and solvent control groups underwent simultaneous treatment for 24 h. Treatment groups were incubated with palmitate at a concentration of 0.25 mM^25,35,36. For the signal pathway study, cells were pretreated with NAC (N-Acetyl-L-cysteine, a potent Nrf2 agonist) at a concentration of 1 mM for 1 h before being treated with 0.25 mM palmitate^37,38.

Lentivirus infection

Lentivirus infection of RIN-m5f cells was performed as previously described²⁵. To establish stable AQP7 overexpression in RIN-m5f cells, the AQP7-EGFP-puromycin lentivirus, encoding rat AQP7 gene, was transduced into cells. In parallel, the EGFP-puromycin lentivirus, functioning as empty control. For AQP7 knockdown, the AQP7-gcGFP-puromycin lentivirus, which carries the target sequence (5′-CTGCAGCTACCACCTACTTAA-3′) was utilized. The negative control lentivirus, gcGFP-puromycin, contained the sequence (5′-TTCTCCGAACGTGTCACGT-3′). After digesting and counting cells, dilute them to 1 × 10⁵ cells/ml. Seed 100 µl/well (1*10⁴ cells) in a 96-well plate, and incubate overnight in a 37 °C, 5% CO₂ incubator. After removing the original culture medium, add 50 µl of fresh culture medium to each well. Based on the determined MOI, add an appropriate volume of virus for infection. After 4 h of infection, supplement the volume to 100 µl. After 24 h, remove the culture medium containing the virus and replace it with fresh complete medium, then continue incubation at 37 °C. Clones with sustained AQP7 expression were chosen after a 2-week puromycin (0.5–2 µg/ml) selection period. GeneChem Co., Ltd. (China) performed the vector construction, which was verified by Sanger sequencing, as well as the virus packaging and viral supernatant collection.

Luciferase reporter assay

It was predicted using the Targetscan (version 8.0, https://www.targetscan.org/) that the mRNA fragment of AQP7 contains a binding site for miR-3584-5p. 293T cells (1*10⁶ cells) were co-transfected in 12-well plate for 6 hours with the AQP7-WT and AQP7-MUT vectors (GenePharma, SuzhouChina) (1 µg/well), miR-3584-5p mimic, and NC mimic (0.1 nmol/well). Following transfection, the culture medium was substituted with fresh medium, and the cells were incubated for 48 h at 37 °C with 5% CO₂. Luciferase activity was subsequently assessed utilizing a Dual-Luciferase Reporter Assay Kit (GenePharma, Suzhou, China), in accordance with the manufacturer’s protocol. Luminescence was measured with a microplate reader (Thermo, Varioskan LUX). The experiment was performed in 3 independent biological replicates.

Lipid peroxidation detection

To assess the level of lipid peroxidation in RIN-m5f cells, BODIPY 581/591 C11 (S0043S, beyotime) was used. Initially, RIN-m5f cells were cultivated at 37 °C with 5% CO₂ in 6-well plates. Subsequently, cells were washed in sterile 0.01 M PBS, after which excess PBS was removed. Subsequently, 1 ml of BODIPY 581/591 C11 stain solution was introduced to every well. Then the samples were incubated at 37 °C with 5% CO₂ for 30 min. Upon completion of the incubation period, the supernatant was aspirated once more, and the well plate was washed twice with PBS to eliminate any residual material. Finally, 1 ml of PBS was added to each well, and the samples were analyzed using a fluorescence microscope (IX73, Olympus).

Fe²⁺ detection

To assess the intracellular Fe²⁺ concentration in RIN-m5f cells, a Fe²⁺ probe (FerroOrange- F374, DOJINDO, Japan) was used. The cells were incubated overnight at 37 °C with 5% CO₂ incubator. The supernatant was then discarded, and the cells were washed 3 times with PBS. The FerroOrange working solution, modified to a final content of 1 µmol/L, was then introduced to every well and maintained for 30 min under identical conditions. Finally, the distribution of intracellular Fe²⁺ was assessed using a fluorescence microscope (IX73, Olympus) without the need for additional washing steps.

Western blot analysis

RIN-m5f cells were lysed, and the protein contents of the resultant homogenates were quantified using a BCA protein assay kit (Beyotime Institute of Biotechnology). A total of 50 µg of protein was subjected to separation via 10% SDS-PAGE, followed by transfer onto nitroate cellulose sheets (NC). Next, the NC was blocked with 5% skimmed milk. Following blocking, the NC were incubated with primary antibodies against ACSL4 (T510198S, 1:2000, rabbit anti-ACSL4, Abmart), GPX4 (T56959S, 1:2000, rabbit anti-GPX4, Abmart), AQP7 (SC376407, 1:500, mouse anti-AQP7, Santa Cruz), Nrf2 (AF0639, 1:2000, rabbit anti-Nrf2, Affinity), HO-1 (AF5393, 1:2000, rabbit anti-HO-1, Affinity), β-actin (AC006, 1:5000, rabbit anti-β-actin, ABclonal) or α-tubulin (AC012, 1:5000, mouse anti-α-tubulin, ABclonal). After the incubation, the NC were treated with HRP-linked Goat Anti-Rabbit IgG (E030120, 1:10000, Earthox), HRP-linked Goat Anti-Mouse IgG (E030110, 1:10000, Earthox). Protein bands were detected. Immunoblots were performed using ECL kits (RM02867, ABclonal) and analyzed by ImageJ (version 1.54; National Institutes of Health, USA; https://imagej.net/ij/). The data were standardized, and each experiment was conducted in triplicate.

Reverse transcription quantitative real-time polymerase chain reaction (RT-qPCR)

Total miRNAs were isolated from the RIN-m5f cells using RNAiso Plus reagent (TaKaRa, Japan). RT-qPCR was conducted with the PrimeScriptTM RT Reagent kit and Green^® Premix Ex TaqTM II (Tli RNaseH Plus) kit (TaKaRa, Japan). The expression levels of miR-3584-5p, miR-296-3p, and let-7a-2-3p were quantified using the 2 − ΔΔct method. The sequences of the RT-PCR primers are provided in Table 1.

Table 1.

The sequence of primers used in RT-qPCR.

Names		Sequences
rno-miR-3584-5p	Forward	CGCGGGGAGGAGTCCA
	Reverse	GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACGCCTCC
rno-miR-296-3p	Forward	CGGAGGGTTGGGTGGAGG
	Reverse	GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACGGAGAG
rno-let-7a-2-3p	Forward	CGCGCTGTACAGCCTCCTAG
	Reverse	GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACGGAAAG
U6	Forward	CTCGCTTCGGCAGCACA
	Reverse	AACGCTTCACGAATTTGCGT

GSH analysis

The GSH levels were analyzed using the Micro Reduced Glutathione (GSH) Assay Kit (KTB1050, Abbkine, China). Cell samples (5*10⁶ cells) were collected and washed with PBS, followed by centrifugation at 800 g for 2 min, and the supernatant was discarded. 1 ml of Extraction Buffer was added, and the cells were sonicated on ice for 5 min (30% power, 3 s on, 7 s off, continuous for 5 min). The sample was then centrifuged at 8000 g at 4 °C for 10 min, and the supernatant was collected and kept on ice for subsequent measurement. The detection system was set up according to the instructions provided by the assay kit to prepare blank, standard, and sample wells, ensuring thorough mixing. The reaction mixture was incubated in the dark at room temperature for 2 min. The absorbance at 412 nm was measured using a microplate reader (Multiskan™ FC, Thermo Fisher Scientific, USA) with the SkanIt (version 7.1; Thermo Fisher Scientific, Waltham, MA, USA; https://www.thermofisher.com/), and the GSH content in the sample was calculated.

MDA analysis

The MDA levels were determined using the Micro Lipid Peroxidation (MDA) Assay Kit (KTB1600, Abbkine, China). Cell samples (1*10⁷ cells) were collected and washed with PBS. After centrifugation, the supernatant was discarded. The cells were resuspended in 0.5 ml of pre-chilled Extraction Buffer and sonicated on ice for 5 min (20% power, 3 s on, 7 seconds off, continuous for 5 min). The sample was then centrifuged at 13,000 g at 4 °C for 10 min, and the supernatant was collected and kept on ice for subsequent analysis. The reaction system was prepared according to the instructions provided by the assay kit, thoroughly mixed, and incubated in a 95 °C water bath for 30 min. The sample was then cooled on ice and centrifuged at 10,000 g at 25 °C for 10 min. A 200 µl aliquot of the supernatant was transferred to a 96-well plate, and absorbance values were measured at 532 nm and 600 nm using a microplate reader (Multiskan™ FC, Thermo Fisher Scientific, USA) with the SkanIt (version 7.1; Thermo Fisher Scientific, Waltham, MA, USA; https://www.thermofisher.com/). The MDA content was then calculated.

H₂O₂ analysis

Cellular H₂O₂ content was measured using the Micro Hydrogen Peroxide (H₂O₂) Assay Kit (KTB1041, Abbkine, China). A total of 5 *10⁶ cells were collected, washed with PBS, and then resuspended in 1 ml of pre-chilled Assay Buffer. The cells were homogenized on ice, followed by centrifugation at 12,000 g at 4 °C for 5 min. The supernatant was collected for further analysis. The reaction system was prepared according to the instructions provided in the assay kit, thoroughly mixed, and incubated at 37 °C for 10 min. The absorbance was measured at 580 nm using a microplate reader (Multiskan™ FC, Thermo Fisher Scientific, USA) with the SkanIt (version 7.1; Thermo Fisher Scientific, Waltham, MA, USA; https://www.thermofisher.com/). The intracellular H₂O₂ content was then calculated based on the standard curve.

Statistical analysis

All data are presented as the mean ± SEM, with statistical analysis performed using student’s t-test or analysis of variance (ANOVA) with GraphPad Prism (version 9.5.0, San Diego, CA, USA; https://www.graphpad.com). p ≤ 0.05 was considered statistically significant.

Results

Under high-lipid condition, the expression of AQP7 in RIN-m5f islet β-cells was decreased, oxidative stress was activated, and ferroptosis occurred

Numerous investigations have shown that elevated levels of the palmitic acid are associated with β-cell dysfunction³⁹. Extended exposure to increased levels of free fatty acids may result in cellular injury or apoptosis⁴⁰. To detect changes in islet β-cells under high-lipid conditions, we integrated and adjusted other research models^41,42. AQP7 protein expression in RIN-m5f cells was significantly reduced under lipid-rich conditions, compared to the control group (Ctrl). Additionally, the intracellular H₂O₂ content was increased (Fig. 1A-C). Additionally, to examine the alterations in oxidative stress within islet β-cells under lipid-rich circumstances, the expression of HO-1 and Nrf2 proteins was markedly diminished (Fig. 1D-F), the intracellular MDA level was elevated, and the intracellular GSH content was reduced (Fig. 1G-H). To investigate ferroptosis in Islet β-cells under high-lipid conditions, we detected increased expression of acyl-coa synthase (ACSL4) and decreased expression of glutathione peroxidase 4(GPX4) (Fig. 2A-C). In addition, intracellular lipid peroxidation levels increased and intracellular Fe²⁺ accumulated (Fig. 2D-G).

Fig. 1.

Effect of palmitic acid on AQP7 and oxidative stress-related pathway proteins expression and H₂O₂, MDA, GSH concentration detected by Western blotting and Elisa. (A) the expression of AQP7 in control (Ctrl), 0.25 mM sodium palmitate (PA) treated groups. (B) graph indicates the ratio of AQP7/α-tubulin (n = 3). (C) graph indicates intracellular H₂O₂ content (n = 3). (D) the expression of Nrf2, HO-1 in control (Ctrl), 0.25 mM sodium palmitate (PA) treated groups. (E) graph depicts Nrf2/α-tubulin ratio (n = 3). (F) graph depicts HO-1/α-tubulin ratio (n = 3). (G) graph indicates intracellular MDA content (n = 3). (H) graph indicates intracellular GSH content (n = 3). * P ≤ 0.05.

Fig. 2.

Impact of palmitic acid on ferroptosis. (A) the expression of ACSL4, GPX4 in control (Ctrl), 0.25 mM sodium palmitate (PA) treated groups. (B) graph illustrates ACSL4/β-actin ratio (n = 3). (C) graph illustrates GPX4/β-actin ratio (n = 3). (D) BODIPY 581/591 was applied to detect changes in lipid peroxidation in control (CTRL), 0.25 mM sodium palmitate (PA)-treated groups. (E) graphs display the ratio of fluorescence intensity of oxidized lipids/unoxidized lipids (n = 3). (F) the content of intracellular Fe²⁺ in the control group (CTRL) and 0.25 mM sodium palmitate (PA) treated groups was detected by fluorescence probe. (G) The quantitative analysis of intracellular Fe²⁺ fluorescence intensity is presented in the graph (n = 3). * P ≤ 0.05.

AQP7 regulates oxidative stress and ferroptosis

To investigate whether oxidative stress and ferroptosis are regulated by AQP7 under high-lipid conditions, we successfully established RIN-m5f cell lines stably knockdown or overexpressing AQP7 under lentivirus infection, as shown in (S1 Fig. 1). Subsequently, we investigated RIN-m5f cells that were knocked down AQP7(KD) and added the Nrf2 agonist NAC, and the results suggested that knockdown of AQP7 resulted in H₂O₂ accumulation in RIN-m5f cells (Fig. 3A-C), decreased expression of Nrf2 and HO-1 (Fig. 3D-F), activation of oxidative stress (Fig. 3G-H). The expression of GPX4 was notably reduced, while the expression of ACSL4 showed a significant increase (Fig. 4A-C), the level of lipid peroxidation was increased (Fig. 4D-E), and the total amount of intracellular Fe²⁺ was increased (Fig. 4F-G). However, NAC application alleviated the above changes caused by AQP7 knockdown.

Fig. 3.

AQP7 knockdown promotes H₂O₂ accumulation and oxidative stress. (A) the expression of AQP7 in the control group (KDc), AQP7 knockdown group (KD), AQP7 knockdown and applied Nrf2 agonist group (KD + NAC). (B) graph indicates the ratio of AQP7/α-tubulin (n = 3). (C) graph indicates intracellular H₂O₂ content (n = 3). (D) the expression of Nrf2, HO-1 in control group (KDc), AQP7 knockdown group (KD), AQP7 knockdown and application of Nrf2 agonist group (KD + NAC). (E) graph illustrates Nrf2/α-tubulin ratio (n = 3). (F) graph illustrates HO-1/α-tubulin ratio (n = 3). (G) graph indicates intracellular MDA content (n = 3). H: graph indicates intracellular GSH content (n = 3). * P ≤ 0.05.

Fig. 4.

Impact of AQP7 knockdown on ferroptosis. (A) the expression of ACSL4, GPX4 in control group (KDc), AQP7 knockdown group (KD), AQP7 knockdown and application of Nrf2 agonist group (KD + NAC). (B) graph represents ACSL4/β-actin ratio (n = 3). (C) graph represents GPX4/β-actin ratio (n = 3). (D) BODIPY 581/591 was applied to detect changes in lipid peroxidation in different groups. (E) graphs illustrate the ratio of fluorescence intensity of oxidized lipids/unoxidized lipids (n = 3). (F) the intracellular Fe²⁺ concentration in different groups. (G) the quantitative analysis of fluorescence intensity of intracellular Fe²⁺ was shown by graph (n = 3). * P ≤ 0.05.

In addition, overexpression of AQP7 significantly alleviated the accumulation of intracellular H₂O₂ induced by PA (Fig. 5A-C). Moreover, the protein levels of Nrf2 and HO-1 were increased (Fig. 5D-F). The intracellular MDA level decreased, while the total GSH level increased (Fig. 5G-H). As a result, cellular oxidative stress was relieved. In addition, cell ferroptosis was also effectively alleviated. Intracellular ACSL4 levels were reduced, while GPX4 levels were elevated (Fig. 6A-C). Lipid peroxidation levels were diminished (Fig. 6D-E), and Fe²⁺ levels were lowered (Fig. 6F-H).

Fig. 5.

Overexpression of AQP7 prevents H₂O₂ accumulation and oxidative stress. (A) the expression of AQP7 in control group (OEc), AQP7 overexpression group (OE), high-lipid group (PA), AQP7 overexpression group under high-lipid condition (OE + PA). (B) graph illustrates the ratio of AQP7/α-tubulin (n = 3). (C) graph indicates intracellular H₂O₂ content (n = 3). (D) the expression of Nrf2 and HO-1 in different groups. (E) graph illustrates the ratio of Nrf2/α-tubulin (n = 3). (F) graph represents the ratio of HO-1/α-tubulin (n = 3). (G) graph indicates intracellular MDA content (n = 3). (H) graph indicates intracellular GSH content (n = 3). * P ≤ 0.05.

Fig. 6.

Impact of AQP7 overexpression on ferroptosis. (A) the expression levels of ACSL4 and GPX4 in the control group (OEc), AQP7 overexpression group (OE), high-lipid group (PA), and AQP7 overexpression group under high-lipid condition (OE + PA). (B) graph represents ACSL4/β-actin ratio (n = 3). (C) graph represents GPX4/β-actin ratio (n = 3). (D) BODIPY 581/591 was employed to assess alterations in lipid peroxidation across several groups. (E) graph illustrates fluorescence intensity of oxidized lipids/unoxidized lipids (n = 3). (F) the content of intracellular Fe²⁺ in different groups. (G) graph indicates fluorescence intensity quantitative analysis of intracellular Fe²⁺ (n = 3). * P ≤ 0.05.

AQP7 regulates the Nrf2/HO-1 pathway

To investigate whether changes in AQP7 expression affect oxidative stress levels in cells by regulating the Nrf2/HO-1 pathway, we applied NAC to inhibit high-lipid-induced oxidative stress and observed the expression changes of AQP7 at the same time. It was found that NAC did not alleviate the reduced expression of AQP7 under high-lipid conditions (Fig. 7A-C), but significantly inhibited the occurrence of oxidative stress (Fig. 7D-H) and intracellular ACSL4 levels decreased, GPX4 levels increased (Fig. 8A-C). Compared to the PA group, the use of NAC under the PA environment also led to a partial decrease in intracellular lipid peroxidation levels (Fig. 8D-E), and a significant reduction in intracellular Fe²⁺ levels (Fig. 8F-G).

Fig. 7.

Effect of NAC on high-lipid-induced RIN-m5f cells. (A) the expression of AQP7 in control group (CTRL), NAC group (NAC), high-lipid group (PA), NAC applied under high-lipid condition (PA + NAC). (B) graph represents the ratio of AQP7/α-tubulin (n = 3). (C) graph indicates intracellular H₂O₂ content (n = 3). (D) The expression levels of Nrf2 and HO-1 across several groups. (E) graph illustrates Nrf2/α-tubulin ratio (n = 3). (F) graph illustrates HO-1/α-tubulin ratio (n = 3). (G) graph indicates intracellular MDA content (n = 3). H: graph indicates intracellular GSH content (n = 3). * P ≤ 0.05.

Fig. 8.

Impact of NAC on ferroptosis. (A) the expressions of ACSL4, GPX4 in control group (CTRL), NAC group (NAC), high-lipid group (PA), and NAC applied under high-lipid condition (PA + NAC). (B) graph represents ACSL4/β-actin ratio (n = 3). (C) graph represents GPX4/β-actin ratio (n = 3). (D) BODIPY 581/591 was employed to assess alterations in lipid peroxidation within the control group (CTRL), NAC group (NAC), high-lipid group (PA) and NAC group (PA + NAC) under high-lipid condition. (E) graph illustrates the ratio of fluorescence intensity of oxidized lipids/unoxidized lipids (n = 3). (F) the content of intracellular Fe^{2 +} in different groups. (G) graph indicates fluorescence intensity quantitative analysis of intracellular Fe^{2 +} (n = 3). * P ≤ 0.05.

miR-3584-5p targets AQP7

MicroRNAs have been demonstrated to govern essential processes in cell biology and play a pivotal part in the response of many cell types^43–45. In order to predict miRNAs that can target AQP7 mRNA, three possible miRNAs, miR-3584-5p, miR − 296-3p and miR − 7a-2-3p, were obtained from the miRNA database Miranda, Targetscan and miRwalk (Fig. 9A). The above 3 possible miRNA expression levels were examined in RIN-m5f cells cultured under high-lipid conditions, where only miR-3584-5p expression was significantly increased (Fig. 9B-D).

Fig. 9.

Predict and validate the targeting of AQP7 mRNA 3’ UTR region by miR-3584-5p. (A) Miranda, Targetscan, and miRwalk jointly predict mirRNA that bind to the 3’ UTR of AQP7 mRNA. (B) graph represents miR-3584-5p expression changes (n = 3). (C) graph represents miR-296-3p expression changes (n = 3). (D) graphs represent miR-7a-2-3p expression changes (n = 3). (E) Relative luciferase activity of the wild-type or mutant AQP7 reporter in 293T cells co-transfected with miR-3584-5p mimic(miR-3584-5p) or control mimic (miR-Ctrl) (n = 3). (F) possible binding sites of the 3’ UTR region of AQP7 mRNA to miR-3584-5p. * P ≤ 0.05.

Dual-luciferase reporter assay initially validated the binding of miR-3584-5p to AQP7 mRNA (Fig. 9E). Through the target gene prediction site Targetscan, the possible binding site of AQP7 mRNA 3′ UTR region with miR-3584-5p was speculated (Fig. 9F).

miR-3584-5p regulates ferroptosis and AQP7 in high-lipid induced RIN-m5f cells

To investigate if miR-3584-5p can modulate the expression of AQP7 in RIN-m5f cells, miR-3584-5p mimics or inhibitors were administered to RIN-m5f cells. The findings indicated that the introduction of the miR-3584-5p mimic diminished the production of AQP7 and the intracellular buildup of H₂O₂ in RIN-m5f cells (Fig. 10A-D), inhibited Nrf2/HO-1, activated oxidative stress (Fig. 10E-I), and increased lipid peroxidation, intracellular Fe^{2 +} accumulates and ferroptosis is activated (Fig. 11). Whereas after the application of miR-3584-5p inhibitor, ferroptosis is inhibited, the reduction of AQP7 in RIN-m5f cells caused by high-lipid conditions was alleviated (Fig. 12A-D), the inhibition of the Nrf2/HO-1 pathway was weakened (Fig. 12E-G), the intracellular MDA level was reduced, and the GSH content was increased (Fig. 12H-I). In addition, the expression of ferroptosis markers, lipid peroxidation and intracellular Fe²⁺ content demonstrated that inhibition of intracellular miR-3584-5p expression could alleviate cell ferroptosis under high-lipid condition (Fig. 13).

Fig. 10.

miR-3584-5p mimics regulate AQP7 in RIN-m5f cells. (A) AQP7 expression in the control group (Ctrl), miRNA mimic blank control group (mimicCTRL), and miRNA mimic group (mimic). (B) graph indicates the ratio of AQP7/α-tubulin (n = 3). (C) graph indicates intracellular H₂O₂ content (n = 3). (D) graph shows the expression levels of miR-3584-5p in each group after application of miRNA mimics (n = 3). (E) The expression levels of Nrf2 and HO-1 across several groups. (F) graph illustrates Nrf2/α-tubulin ratio (n = 3). (G) graph illustrates HO-1/α-tubulin ratio (n = 3). (H) graph represents intracellular MDA content (n = 3). (I) graph indicates intracellular GSH content (n = 3). * P ≤ 0.05.

Fig. 11.

miR-3584-5p mimics regulate ferroptosis in RIN-m5f cells. (A) the expression of ACSL4 and GPX4 in the control group (CTRL), miRNA mimic blank control group (mimicCTRL), and miRNA mimic group (mimic). (B) graph represents ACSL4/β-actin ratio (n = 3). (C) graph represents GPX4/β-actin ratio (n = 3). (D) BODIPY 581/591 was applied to detect lipid peroxidation changes in different groups. (E) graphs illustrate the ratio of fluorescence intensity of oxidized lipids/unoxidized lipids (n = 3). (F) the intracellular Fe²⁺ concentration in different groups. (G) the fluorescence intensity of intracellular Fe²⁺ (n = 3). * P ≤ 0.05.

Fig. 12.

miR-3584-5p inhibitors regulate AQP7 in high-lipid-induced RIN-m5f cells. (A) the expression level of AQP7 in control group (Ctrl), high-lipid group (PA), blank control group (PA + inhibitorCTRL) and high-lipid group (PA + inhibitor). (B) graph indicates the ratio of AQP7/α-tubulin (n = 3). (C) graph indicates intracellular H₂O₂ content (n = 3). (D) graph indicates miR-3584-5p expression after the addition of miRNA inhibitors to each group (n = 3). (E) the expression of Nrf2 and HO-1 in different groups. (F) graph illustrates Nrf2/α-tubulin ratio (n = 3). (G) graph represents HO-1/α-tubulin ratio (n = 3). (H) graph illustrates intracellular MDA content (n = 3). I: graph indicates intracellular GSH content (n = 3). * P ≤ 0.05.

Fig. 13.

miR-3584-5p inhibitors regulate ferroptosis in high-lipid-induced RIN-m5f cells. (A) the expression of ACSL4 and GPX4 in control group (CTRL), high-lipid group (PA), blank control group (PA + inhibitorCTRL) and high-lipid group (PA + inhibitor). (B) graph represents ACSL4/β-actin ratio (n = 3). (C) graph represents GPX4/β-actin ratio (n = 3). (D) BODIPY 581/591 was applied to detect changes in lipid peroxidation in different groups. (E) graph illustrates fluorescence intensity of oxidized lipids/unoxidized lipids (n = 3). (F) intracellular Fe²⁺ content changes in different groups were detected by applying Fe²⁺ fluorescent probe. (G) graph indicates fluorescence intensity quantitative analysis of intracellular Fe²⁺ (n = 3). * P ≤ 0.05.

Discussion

Our investigation offers significant contributions to understanding the function of AQP7 in oxidative stress and ferroptosis induced by elevated lipid levels. Our findings reveal that under high-lipid conditions, AQP7 expression in RIN-m5f cells is reduced, resulting in the suppression of the Nrf2/HO-1 pathway and the induction of oxidative stress, and subsequent ferroptosis. Additionally, our results indicate that high-lipid conditions upregulate miR-3584-5p, which inhibits AQP7 expression, thereby activating oxidative stress and ferroptosis. Inhibition of miR-3584-5p counteracts the downregulation of AQP7 expression induced by high-lipid exposure, subsequently alleviating oxidative stress and ferroptosis in RIN-m5f cells (Fig. 14). These findings suggest a potential therapeutic approach targeting AQP7 and miR-3584-5p for managing high-lipid induced cellular damage.

Fig. 14.

Summary for regulations of miR-3584-5p and AQP7 on RIN-m5f cells under high palmitic acid (PA) condition.

Chronic hyperlipidemia can induce various metabolic disorders^46–48. Palmitic acid, as a major component of free fatty acids (FFAs), is widely used in studies of lipid-induced damage^49–51 and is the most common substance in research on high-fat-induced β-cell injury^52–54. Palmitic acid has been shown to induce oxidative stress and ferroptosis in β-cells^54,55, which aligns with our findings. In our study, we found that palmitic acid stimulates an increase in miR-3584-5p, which in turn inhibits the expression of AQP7 protein. This leads to the accumulation of intracellular H₂O₂, triggering oxidative stress and ferroptosis in the cells.

Islet β-cell injury significantly contributes to hyperlipidemia in vivo¹⁸. Research has shown the pivotal role of oxidative stress in β-cell impairment and apoptosis¹⁹.The present study has demonstrated that β-cell damage produced by oxidative stress can be substantially alleviated with the use of antioxidants, free radical scavengers, or by overexpressing antioxidant enzymes in pancreatic islets or transgenic mice⁵⁶.

H₂O₂ represents a key reactive oxygen species engaged in the redox control of diverse biological activities. Owing to the restricted expression of mitochondrial and peroxisomal antioxidant enzymes in islet β-cells, the availability of rapidly reactive, high-affinity sites, particularly sulfhydryl groups, is considerably low. This deficiency may compromise the cells’ ability to effectively manage oxidative stress⁵⁷. Thus, only a small fraction of the high concentrations of H₂O₂ produced in mitochondria and peroxisomes is rapidly inactivated at the site of its production, which provides an ideal prerequisite for long-range transmission of H₂O₂. Thus, this indicates that nearly any intracellular macromolecule—be it carbohydrate, lipid, or nucleic acid—may act as a potential target for the deleterious effects of hydroxyl radicals⁵⁸. The passive diffusion of H₂O₂ across biofilms requires specific subtypes of aquaporins⁵⁹. The present study has shown that AQP7 in BMSC can regulate cell proliferation by transporting intracellular H₂O₂²⁴. Previous research has demonstrated that AQP7 deficiency is linked to a decrease in both the number of islet cells and the overall mass of β-cells. In addition, the AQP7 gene in the human genome region has been reported to be associated with the pathogenesis of type 2 diabetes⁶⁰. In our previous study on diabetic rats and INS-1 cells, we observed a notable decrease in AQP7 expression²⁵. In this study, we found that knockdown of AQP7 or reduction of AQP7 induced by high-lipid conditions could increase the accumulation of H₂O₂ in β-cells. AQP7 is believed to serve a vital regulatory role in the oxidative stress and ferroptosis of pancreatic β-cells.

A growing amount of evidence indicates that Nrf2 activation during hyperglycemia, together with the resultant generation of ROS, is crucial for the protection and preservation of β-cell mass. Mice exhibiting worldwide Nrf2 loss demonstrate diminished islet cell viability, reduced islet size, and heightened vulnerability to arsenic-induced β-cell injury, but paradoxically, they also demonstrate increased survival. These findings suggest that Nrf2 deficiency not only increases vulnerability to arsenic-induced damage but also potentially triggers compensatory mechanisms that promote β-cell mass expansion⁶¹, the increase of Nrf2 expression has been demonstrated to augment the proliferation of rat β-cells, along with primary mouse and human β-cells^62,63. A marked reduction in the expression of multiple Nrf2 target genes has been noted in the retinal and dermal tissues of diabetic patients, as well as in the fibroblasts of a diabetic rat model, indicating that chronic in vivo exposure to elevated lipid and glucose levels may significantly contribute to the pathogenesis of diabetes⁶⁴. Our results align with previous reports demonstrating that high fat promotes Nrf2 degradation and increased oxidative stress⁶⁵. This study confirmed that the inhibition of AQP7 under high-lipid conditions down-regulated Nrf2 expression, while overexpression of AQP7 could alleviate the decrease in Nrf2 due to high-lipid. However, whether AQP7 directly affects Nrf2 expression or indirectly through other signaling pathways needs further study. Additionally, multiple studies have indicated that the inhibition of the Nrf2/HO-1 pathway is associated with ferroptosis in conditions such as diabetic nephropathy⁶⁶, transplanted islets⁶⁷, acute lung injury⁶⁸, and blood-brain barrier dysfunction⁶⁹. Our findings also validate the critical role of the Nrf2/HO-1 pathway in ferroptosis. Furthermore, our study confirms that the activation of Nrf2/HO-1 in β-cells can inhibit ferroptosis and reduce the accumulation of intracellular Fe²⁺.

MiR-3584-5p has been associated with the regulation of multiple biological processes, including inflammation^70,71, proliferation⁷², pyroptosis⁷³, there is limited research on the effect of miR-3584-5p following high-lipid induced injury in pancreas β-cells. In this study, our findings indicated a significant elevation in the expression of miR-3584-5p after high-lipid-induced injury. The expression of AQP7 decreased by miR-3584-5p mimics and increased by miR-3584-5p inhibitors compared with high-lipid injury cells. Additionally, lipid peroxidation and ferroptosis-related assays showed that miR-3584-5p mimics activated, while inhibitors suppressed ferroptosis markers. Our findings suggest that miR-3584-5p contributes to ferroptosis in high-lipid conditions by modulating AQP7. However, there are some limitations in our study. While we focused on the effect of miR-3584-5p on AQP7, we did not examine other potential mechanisms, such as its impact on transcription⁷⁴ or protein degradation⁷⁵ pathways, that could also influence protein expression. Furthermore, although our results suggest a role for miR-3584-5p in ferroptosis, more detailed studies are needed to explore the exact molecular mechanisms involved.

Targeting ferroptosis in the context of diabetes could represent a potentially novel strategy for the preservation of the β-cell population⁷⁶. While essential, excessive iron levels can be toxic, hence, its concentration must be tightly regulated within an optimal range. Iron homeostasis is meticulously controlled through the processes of iron metabolism⁷⁷. Increasing evidence suggests that iron accumulation may play a part in the pathophysiological mechanisms of β-cell failure^78–80. In our research, the knockdown of AQP7 further exacerbated oxidative stress and ferroptosis, while upregulation of AQP7 alleviated these effects. Activation of the Nrf2 pathway with specific agonists reduced oxidative stress and ferroptosis even in the absence of AQP7, suggesting an important role played by AQP7 in regulating ferroptosis in pancreatic β-cells.

In high PA condition, miR-3584-5p upregulation inhibits AQP7 expression, impairing H₂O₂ transport and suppressing the Nrf2/HO-1 pathway, thereby promoting oxidative stress and ferroptosis in RIN-m5f cells.

Limitations and future perspectives

Although this study has established an initial understanding of the miR-3584-5p/AQP7 pathway under high-lipid conditions, it is important to recognize some limitations. The mechanisms that govern miR-3584-5p upregulation under high-lipid conditions deserve further investigation. Furthermore, the contribution of the miR-3584-5p/AQP7 axis to the progression of pancreatic β-cell injury under high-lipid conditions and its therapeutic potential should be further evaluated in animal models.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Ce Luan contributed to conceptualization, methodology, formal analysis, investigation, data curation, visualization, and wrote the original draft. Zhi Wang and Meijie Li conducted formal analysis and provided supervision. Fei Gao contributed to conceptualization, secured funding, supervised the work, and reviewed and edited the manuscript. Ruixi Feng and Yishou Wang developed the methodology and provided supervision. Junjie Wu and Shu Yang contributed to conceptualization and project administration. As the corresponding author, Mei Yang led the conceptualization, secured funding, developed the methodology, supervised the project, and was responsible for writing the original draft as well as reviewing and editing the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China to Mei Yang (No. 81971230 and 81671312) and Fei Gao (No. 82370691), Natural Science Foundation of Chongqing, China to Mei Yang (No. CSTB2023NSCQ- MSX0565) and Fei Gao (No. CSTB2023NSCQ-MSX0072).

Data availability

All data generated or analysed during this study are included in this published article and its supplementary information files.

Declarations

Competing interests

The authors declare no competing interests.