AMSC-EVs attenuate acute kidney injury through TXNIP-IKKα/NFκB signaling-mediated renal CX3CR1⁺ macrophage polarization

Abstract

Background

Extracellular vesicles generated from mesenchymal stem cells (MSC-EVs) have garnered significant attention as a cell-free treatment option for acute kidney injury (AKI). The fundamental processes and capabilities of MSC-EVs in attenuating kidney injury are still largely unclear.

Methods

The AKI mouse model was established by intraperitoneal injection of cisplatin. AKI mice were further randomized to receive Phosphate- Buffered Saline, adipose-derived mesenchymal stem cell-extracellular vesicles (AMSC-EVs (50 µg), or AMSC-EVs (100 µg) into the tail vein. Following a 96-hour post-injury period, the mice were euthanized, and kidney tissues together with blood samples were procured for paraffin embedding and immunoblotting. To investigate the relationship between AMSC-EVs and renal CX3CR1⁺ macrophages, CX3CR1⁺ macrophage-specific conditional knockout mice (CX3CR1-Cre+/-; Rosa26-LSL-DTR+/-) were generated. Additionally, overexpression of experiments of thioredoxin-interacting protein (TXNIP) were conducted to analyze macrophage polarization and TXNIP-IKKα/NFκB signaling pathway expression.

Results

In mice, AMSC-EVs reduced the renal tubule damage and ameliorated cisplatin-induced AKI in a dose-dependent manner. However, in the CX3CR1⁺ macrophage ablated group, AKI mice exhibited more severe renal tubular pathology compared to littermate controls, suggesting diminished therapeutic efficacy of AMSC-EVs post CX3CR1⁺ macrophage ablation. Meanwhile, AMSC-EVs promoted polarization of renal CX3CR1⁺ macrophages towards reparative M2 macrophages, leading to increased production of anti-inflammatory factors and subsequent alteration of the inflammatory microenvironment in renal tubular cells, thereby facilitating the self-repair process in AKI mice. Mechanistically, AMSC-EVs suppressed the protein expression of TXNIP-IKKα/NFκB in renal CX3CR1⁺ macrophages. Finally, overexpression of TXNIP appeared to attenuate the protective effects of AMSC-EVs in renal CX3CR1⁺ macrophages.

Conclusions

Our study findings suggest that AMSC-EVs modulate the polarization of renal CX3CR1⁺ macrophages and promote renal self-recovery following cisplatin-induced AKI through the TXNIP-IKKα/NFκB signaling pathway.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s13287-025-04754-z.

Introduction

Acute kidney injury (AKI) is a clinical syndrome characterized by a rapid decline in renal function over a short period and can arise from diverse etiologies. This global health burden exhibits increasing prevalence in both developed and developing nations, with high associated morbidity and mortality rates Cisplatin represents a significant contributor to clinical AKI, although the precise underlying molecular mechanisms remain incompletely understood [1]. Currently, effective strategies for preventing or treating AKI are limited. Consequently, exploring approaches to mitigate AKI-related damage to vital organs and developing targeted therapeutics hold significant clinical importance for advancing AKI management.

Mesenchymal stem cells (MSCs) have gained widespread use in regenerative medicine due to their multilineage differentiation potential and immunomodulatory properties, making them applicable for treating various diseases, including kidney disorders. MSCs can be isolated from multiple tissues, such as bone marrow (BMMSC) [2], adipose tissue (AMSC) [3], umbilical cord [4], placenta [5], dental pulp, skin, blood, and urine [6]. Research indicates that MSCs migrate to sites of injury primarily via paracrine mechanisms and direct contact with minimal reliance on differentiation capabilities [7]. Following injection, labeled MSCs specifically localize to injured kidney areas [8]. However, the clinical application of MSCs faces limitations due to ethical concerns, safety issues, and sourcing challenges. Extracellular vesicles (EVs) derived from MSCs offer numerous advantages over the cells themselves, positioning them as promising cell-free therapeutics for inflammatory disorders [9, 10]. Emerging evidence suggests that AMSC-derived extracellular vesicles (AMSC-EVs) play a significant role in treating AKI [11], although their exact mechanism of action requires further elucidation.

In recent years, with the rapid development of single-cell RNA sequencing (scRNAseq), more and more literature has emphasized the evolving heterogeneity of macrophages between and within tissues. CX3C motif receptor 1 (CX3CR1), also known as G protein-coupled receptor 13, is the specific receptor for fractalkine (CX3CL1), which can bind to either membrane-bound or soluble CX3CL1 to exert different functions [12]. CX3CR1 ⁺ macrophages have been found to play different roles in a variety of tissues. In the central nervous system, CX3CR1 is considered a specific marker of microglia. Inflammatory responses can stimulate CX3CR1 ⁺ microglia to bind with CX3CL1 released by nerve cells, exerting anti-inflammatory and neuroprotective effects [13–15]. In the cardiovascular system, CX3CR1 ⁺ macrophages are considered to be resident macrophages in the heart. Through their phagocytic function, they can process abnormal mitochondria and exert cardiomyocyte-protective effects [16–18]. In the intestine, CX3CR1^hi macrophages play different roles at different stages of colitis progression [19]. The Alex Yashchenko team used scRNAseq to discover that the kidneys of healthy mice contained four major monocyte subgroups and two major kidney resident macrophage subgroups (KRM). They also identified CX3CR1 as a gene that governs cortex-specific accumulation of Ccr2 ⁺ KRM and showed that CX3CR1 regulates KRM heterogeneity and niche-specific disease progression [20]. In the unilateral ureter ligation (UUO) model, CX3CR1⁺ macrophages activated fibroblasts by secreting IL−1β, induced α-SMA and fibronectin expression, and promoted collagen deposition [21]. In addition, studies have shown that AMSC attenuate rheumatoid arthritis by restoring the CX3CR1 ⁺ synovial lining macrophage barrier [22]. However, few studies have reported the association of AMSC-EVs and renal CX3CR1 ⁺ macrophages in AKI.

Thioredoxin-interacting protein (TXNIP), a member of the α-arrestin protein family, is markedly elevated in human proteinuric nephropathy [23]and serves as a crucial regulator of the NF-κB signaling cascade [24]. Numerous studies demonstrate that TXNIP modulates macrophage inflammation, with elevated TXNIP levels promoting M1 polarization and inhibiting M2 polarization [25–28].

This study investigates the potential therapeutic role of AMSC-EVs in murine AKI. We employed CX3CR1⁺ macrophage-specific conditional knockout mice to examine the relationship between AMSC-EVs and renal CX3CR1⁺ macrophages. Furthermore, we explored the underlying mechanism of AMSC-EVs in vitro, focusing specifically on the TXNIP-IKKα/NF-κB signaling pathway.

Materials and methods

Isolation and characterization of AMSC-EVs

AMSC were isolated from human adipose tissue samples using established methodologies documented in the literature and were grown in standard media (DMEM/F12, Gibco, USA) supplemented with 10% fetal bovine serum (FBS, Gibco, USA) and 1% penicillin/streptomycin (P/S, Gibco, USA) at 37 °C in a 5% CO2 atmosphere [29].The media was altered every 3 days until adherent cells displayed a spindle-shaped morphology. Upon achieving around 90% confluency, cells were collected with 0.25% trypsin (Gibco, USA) and passaged at a ratio of 1:3 or 1:4 for further proliferation. Passages 3 to 8 of the mesenchymal stem cells were used for isolating EVs. The supernatant was collected after 48 h of culturing AMSC in DMEM/F12 media containing 10% EVs-depleted FBS. The debris and dead cells were removed from the medium by two-step centrifugation (1000 g for 15 min and 10,000 g for 30 min at 4◦C). After centrifuging the cell-free supernatant for 80 min at 4 °C at 110,000 g, washing it with PBS, and performing another high-speed centrifugation under the same settings, the mixture was removed. After being cleansed, the AMSC-EVs were again suspended in PBS and kept cold, at -80 °C [30].

Subsequently, the particle size and zeta potential were measured utilizing Nanoparticle Tracking Analysis (NTA, ZetaView PMX 110). The transmission electron microscopy (TEM, HITACHI) was employed to examine the morphological features of the EVs. Additionally, immunoblotting was conducted to detect extracellular vesicles protein markers ((CD9, 1:1000, A19027, ABclonal, Wuhan, China); CD63 (1:1000, A19023, ABclonal, Wuhan, China) and CD81 (1:600, 66866-1-Ig, Proteintech, Wuhan, China)).

Cell culture

Human monocytic-leukemia cells (THP-1, Shanghai Zhong Qiao Xin Zhou Biotechnology Co.,Ltd.) were cultured in RPMI 1640 media (Gibco, USA) containing 10% FBS and 1% P/S. For some experiments, THP-1 cells, exhibiting optimal growth circumstances, were uniformly plated in Petri dishes and subsequently differentiated into M0 macrophages following stimulation with Phorbol 12-Myristate 13-Acetate (PMA) (16561-29-8, MedchemExpress, Shanghai, China) at a concentration of 100 ng/ml for 24 h. M0 macrophages were further treated with 1 ug/mL cisplatin (HY-17394, MedChemExpress, Shanghai, China) for 24 h to establish an in vitro cell injury model.

Cellular uptake of AMSC-EVs in vitro

Fluorescent staining using 1,1′-dioctadecyl-3,3,3′,3′-Tetramethylindocarbocyanine perchlorate (DiI, Invitrogen) was conducted to label AMSC-EVs. The DiI-labeled EVs were co-cultured with THP-1 cells for 6 h in FBS-free 1640 media at 37 °C. The nuclei were stained with DAPI (Sigma, USA) at 37 °C for a duration of 10 min. Following washing, the stained cells were examined microscopically (Olympus, Tokyo, Japan).

Animal models and therapeutic experiments

8–10 week old Nude male mice were purchased from Nanjing Zhonghua Biotechnology Co., Ltd. Rosa26-LSL-DTR strain mice (NM-KI-210094) were purchased from Shanghai Nanfang Model Biotechnology Co., Ltd. CX3CR1-Cre strain mice (T006768) were purchased from Jiangsu Jicui Yaokang Biotechnology Co., Ltd. All experimental animals were housed in a stable environment with a temperature range of 20–22 °C, a humidity range of 50–70%, a light cycle of 12 h/12 h, and a normal supply of food and water. The experimental methods and standard experimental procedures of laboratory animals were approved and supervised by the Ethics Committee for the Use and Management of Laboratory Animals of Nanjing Medical University. Humane endpoints were established to ensure animal welfare, including significant weight loss (> 30%), behavioral changes (e.g., reduced activity), and signs of pain or distress.

Twenty-four mice were randomly assigned to four treatment groups (6 mice per group). A total of 201 mice were included in this study. The experimental unit was a single mouse, and each animal was randomly assigned to one group. The grouping was performed using a computer-generated randomization list. AKI animal models were created with intraperitoneal administration of 12 mg/kg cisplatin. At 0 h and 24 h after intraperitoneal administration of cisplatin, the mice received injections of either phosphate-buffered saline (PBS), AMSC-EVs (50 µg), or AMSC-EVs (100 µg) into the tail vein. The sample size was estimated based on our previous publications, and no a priori power analysis was performed. No inclusion or exclusion criteria were defined, and no animals were excluded from the analysis. At 96 h post-injury, mice were deeply anesthetized via inhalation of 3–4% isoflurane (R510-22-10, Shenzhen RWD Life Science Co., Ltd.) delivered in oxygen at a flow rate of 1 L/min. Anesthesia was maintained for 2–3 min until a surgical plane was achieved, as confirmed by the absence of the righting reflex and a lack of response to a noxious toe pinch. Following confirmation of deep anesthesia, euthanasia was performed by cervical dislocation. Death was verified by the cessation of breathing and the absence of a corneal reflex. We implemented a blinding procedure so that the researchers performing the measurements were unaware of the treatment groups, thereby minimizing observer bias.

In vivo uptake experiments of AMSC-EVs

To label AMSC-EVs in vivo, we intravenously injected PKH26 (Red Fluorescent Cell Linker Kits MINI26, Sigma, USA)-labeled AMSC-EVs into the tail vein of mice. Mice were euthanized 4 hours post tail vein injection. Kidney samples were fixed with 4% paraformaldehyde and then dehydrated with 30% sucrose solution, followed by embedding and sectioning into 6 μm slices. The location of EVs in kidney slices was assessed using immunofluorescence labeling with the EV marker CD81. Finally, by an LSM880 laser scanning confocal microscope (ZEISS, Germany) to observe the colocalization of renal and EVs markers.

Cell transfection

THP-1 cells, which were well-conditioned, were plated in cell culture plates for TXNIP overexpression. Upon reaching 60–80% confluency, THP-1 cells were transfected with pc-DNA-TXNIP (Genechem, Shanghai, China) using Lipofectamine 3000 (Invitrogen, USA) reagent as per the manufacturer’s instructions. Subsequent to the administered treatment, the cells were collected and further analysis was conducted.

Measurement of creatinine and blood urea nitrogen

Blood samples were collected by removing the eyeball under anesthesia. A 10-minute centrifugation at 3500 rpm was performed on blood samples. The supernatant was obtained, and serum creatinine (Scr, C011-2-1) and blood urea nitrogen (BUN, C013-2-1, Nanjing Jiancheng, Nanjing, China) were quantified utilizing the appropriate assay kits.

Histological analysis and immunohistochemistry analysis

After fixation in 4% paraformaldehyde for over 48 hours, the kidney tissues were embedded and cut into 4 μm slices. Hematoxylin and eosin (HE) and periodic acid-Schiff (PAS) were used to stain the sections. For immunohistochemistry (IHC) labeling, slices were stained with anti-NGAL (sc-515876, 1:50, Santa Cruz, USA)and anti-KIM-1 (30948-1-AP, 1:100, Proteintech, Wuhan, China). The slides were stained with the appropriate HRP-conjugated secondary antibody and DAB reagent (ZhongshanJinqiao, Beijing, China) after being incubated with the primary antibody at 4 °C overnight. A microscope from Olympus in Tokyo, Japan, was used to acquire the images.

Reverse transcription quantitative polymerase chain reaction (RT-qPCR)

Total RNA was isolated from kidney tissues or THP-1 cells using TRIzol reagent (16096020, Invitrogen, Shanghai, China). 1 µg of total RNA was reverse transcribed into cDNA with the Evomac-Mlo Reverse Transcription Kit (Accurate Biology, Beijing, China). Quantitative PCR was conducted using ChamQ Universal SYBR RT-qPCR Master Mix (Accurate Biology, Beijing, China). The primers used for cDNA PCR amplification are listed in Table S1.

Western blot

Protein was extracted from kidney tissue or THP-1 cells using RIPA lysis buffer with 1% protease inhibitor, and the protein concentration was quantified using the BCA protein assay (P0011, Beyotime, Shanghai, China). After transfer, the membranes were incubated overnight at 4 °C with the specified primary antibodies: anti-iNOS (AF0199, 1:1000, Affinity), anti-Arg1 (DF6657, 1:1000, Affinity), anti-TXNIP (18243-1-AP, 1:1000, Proteintech, Wuhan, China), anti-IKKα (ab32041, 1:10000, Abcam, Cambridge, MA), anti-P-IKKα (AP0505, 1:1000, ABclonal, Wuhan, China), anti-NF-κB (8242, 1:1000, CST, USA), anti-P-NF-κB (3033, 1:1000, CST, USA), and anti-GAPDH (10494-1-AP, 1:10000, Proteintech, Wuhan, China). Subsequently, the membranes were incubated with the corresponding HRP-conjugated anti-mouse (SA00001-1, 1:10000, Proteintech, Wuhan, China) or anti-rabbit IgG antibodies (SA00001-2, 1:10000, Proteintech, Wuhan, China) at room temperature for 2 hours. The protein bands were visualized using Image-Pro Plus software (version 6.0).

Immunofluorescence staining

In order to conduct immunofluorescence (IF) analysis, tissue slices or THP-1 cells were first incubated with primary antibodies (anti-F4/80 (30325, 1:200, CST, USA), anti-CX3CR1 (DF7095, 1:200, Affinity, USA), anti-TXNIP (AEBA-20, 1:100, Boster, Germany, anti-CD206 (24595, 1:400, CST, USA)) for an overnight period at 4 °C. Then, secondary antibodies were added in a dark room for 2 hours. The nuclei were stained with DAPI for 10 minutes in the dark. The photos were ultimately acquired utilizing a Zeiss fluorescence microscope.

Statistical analysis

All statistical analyses were performed using GraphPad Prism 8.0 and SPSS 22.0. Quantitative data were described as mean ± standard error of mean, and comparisons between the two groups of quantitative data were made using a two-tailed t-test. The disparities across three or more groups were analyzed utilizing one-way analysis of variance (ANOVA) with Dunnett’s adjustment. P < 0.05 is regarded as a statistically significant difference. All experiments were repeated at least 2–3 times.

Results

Characterization and tracer studies of AMSC-EVs

AMSCs were positive for CD105, CD90, and CD73 but negative for CD45, CD34, and HLA-DR (Figure S1A), and exhibited typical spindle-shaped fibroblast morphology(Figure S1B). After 3 weeks of osteogenic induction or 2 weeks of adipogenic induction, the cells’ mesenchymal origin was confirmed by Alizarin Red or Oil Red O staining (Figure S1C and S1D).

NTA (Fig. 1A) and TEM (Fig. 1B) revealed that AMSC-EVs had diameters of 60–120 nm and typical cup-shaped morphology. Western blotting showed strong expression of the EV markers CD9, CD63, and CD81 (Fig. 1C).

Fig. 1.

Characterization and tracer studies of AMSC-EVs. A Analysis of AMSC-EVs nanoparticles. B Analysis of AMSC-EVs via TEM. (Scale bar, 100 nm). C Western blot was conducted to detect EVs markers CD9, CD63 and CD81. D Representative fluorescent images demonstrated the uptake of DiI-labeled AMSC-EVs by THP-1 cells. Scale bar, 20 μm. E The confocal laser microscope micrographs showing that the PKH26-labeled AMSC-EVs are internalized by cisplatin-treated kidney, Scale bar, 50 μm

We tested the uptake of AMSC-EVs by THP-1 cells. After incubating DiI-labeled EVs with THP-1 cells for 6 hours, confocal microscopy showed that the labeled AMSC-EVs were taken up by macrophages (Fig. 1D). We further examined the distribution of AMSC-EVs in cisplatin-treated kidneys. PKH26-labeled AMSC-EVs were injected into the tail vein of mice, which were then euthanized after 4 hours. Red fluorescent PKH26-labeled EVs were observed in the kidneys treated with cisplatin. Immunofluorescence staining with an anti-CD81 antibody showed that a few red fragments colocalized with CD81 (Fig. 1E), suggesting successful delivery of EVs to the kidneys.

AMSC-EVs attenuate cisplatin-induced kidney injury in mice

To assess the therapeutic efficacy of AMSC-EVs in AKI, a murine model of cisplatin-induced AKI was established, and varying dosages of AMSC-EVs (50 or 100 µg) or PBS were administered via the tail vein at 0 h and 24 h (Fig. 2A). Mice administered cisplatin demonstrated markedly increased Scr and BUN levels, extensive edema, necrosis, epithelial cell detachment, and cast formation, all of which were mitigated by AMSC-EVs therapy in a dose-dependent fashion (Fig. 2B–D). Next, to further verify the therapeutic effect of AMSC-EVs, we used IHC staining to detect NGAL and KIM-1, which are indicators of renal tubular damage. The findings indicated that the expression of NGAL and KIM-1 was reduced by treatment with AMSC-EVs (Fig. 2D). Together, these data indicate that AMSC-EVs treatment attenuates cisplatin-induced kidney injury and reduces proximal renal tubular damage in mice.

Fig. 2.

Therapeutic effects of AMSC-EVs on a mouse model of cisplatin-induced AKI. A Schematic representation of the experimental design. Briefly, mice were administered PBS, 50 μg AMSC-EVs, or 100 μg AMSC-EVs via the tail vein at 0 and 24 hours following cisplatin-induced modeling, and were euthanized 96 hours post-model induction. B, C Impact of AMSC-EVs on Scr and BUN (n = 5–6). D Representative renal sections stained with HE, PAS, and tubular injury markers (NGAL, KIM-1), along with tubular injury scores and quantitative immunohistochemical analyses (n = 6). Scale bar, 50 μm. The information is displayed as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Depletion of renal CX3CR1⁺ macrophages diminishes the effectiveness of AMSC-EVs treatment

The main manifestation of kidney injury is tubular epithelial damage caused by immune activation and inflammatory response mediated by macrophages. We investigated how CX3CR1⁺ macrophages contribute to the benefits of AMSC-EVs therapy. We generated transgenic mice with CX3CR1⁺ macrophage-specific expression of the diphtheria toxin receptor (DTR). The overall procedure, as depicted in the diagram, involved mating CX3CR1-Cre+/- mice with Rosa26-LSL-DTR+/- mice to obtain offspring identified as CX3CR1-Cre+/-; Rosa26-LSL-DTR+/- double heterozygous mice, administering diphtheria toxin (DT) to deplete CX3CR1⁺ macrophages, and followed by sample collection and analysis on day 0 (Figure S2A). Flow cytometric analysis of CX3CR1⁺ macrophages (Figure S2B and S2C) revealed a significant reduction in macrophages within renal tissue at day 0 compared to littermate controls (non-double heterozygous mice), indicating the success of the depletion experiment. To clarify the association between AMSC-EVs and CX3CR1⁺ macrophages, cisplatin-AKI modeling was induced on day 0, followed by AMSC-EVs treatment at 0 h and 24 h, respectively. On day 4, blood and renal tissue samples were collected for biochemical testing and renal histopathological staining. The results showed that the AMSC-EVs treatment group significantly increased the levels of BUN and Scr (Fig. 3A and B) after ablation of CX3CR1⁺ macrophages, and the HE staining and renal pathological damage scores also showed that the ablation group was more severely damaged (Fig. 3C and D), compared with the littermate control group, suggesting that CX3CR1⁺ macrophages play an important role in mediating the renal protective effects of AMSC-EVs. The protective effects of AMSC-EVs appear to be largely dependent on their interactions with CX3CR1⁺ macrophages.

Fig. 3.

Depletion of renal CX3CR1+ macrophages diminished the effectiveness of AMSC-EVs treatment. A, B Bar graphs showing BUN and Scr levels in each group of mice. C, D HE, PAS and IHC staining and histopathological injury scores in mice in the depletion and non-depletion groups, with n = 4 in each group, Scale bars, 50 μm, The findings are provided as the mean ± SEM. *P<0.05, **P<0.01, ***P<0.001

AMSC-EVs alleviate kidney inflammation and promote renal CX3CR1⁺ macrophages transforming towards M2-like phenotype in cisplatin-treated mice

Inflammatory responses are essential for the development and self-healing of AKI.AMSC-EVs polarized renal CX3CR1⁺ macrophages toward an M2 phenotype in cisplatin-AKI. RT-qPCR analysis showed that AMSC-EVs upregulated M2 markers (Arg1, IL-10) while downregulating M1 markers (TNF-α, IL-6, iNOS) (Fig. 4A and B). Western blot analysis confirmed increased Arg1 and decreased iNOS protein expression (Fig. 4C). Immunofluorescence revealed that AMSC-EVs reduced CX3CR1⁺iNOS⁺M1 macrophages and increased CX3CR1⁺CD206⁺M2 macrophages in the kidneys compared to PBS controls (Fig. 4D and E). Taken together, these findings suggest that AMSC-EVs promote the polarization of CX3CR1⁺ macrophages toward the M2 phenotype.

Fig. 4.

The effect of AMSC-EVs on CX3CR1⁺ macrophage polarization in cisplatin-treated mice. A RT-qPCR analysis of TNF-α, IL-6, and IL-10 in kidney tissue of cisplatin-treated mice (n = 3–6). B RT-qPCR analysis of Arg1 and iNOS in the renal tissue of cisplatin-treated mice (n = 4–6). C Western blot analysis of proteins associated with macrophage polarization (n = 3). D, E Immunofluorescence staining showingiNOS and CD206 expression in CX3CR1⁺ macrophages (n = 6), Scale bar, 50 μm. The mean ± SEM is used to express data, *P<0.05, **P<0.01, ***P<0.001

AMSC-EVs convert inflammatory macrophages to an M2 phenotype in vitro

THP-1, an established immortalized human monocyte-like cell line, constitutively expresses CX3CR1 and is widely used for in vitro studies of monocyte differentiation and macrophage function [31, 32]. To better understand how AMSC-EVs interact with CX3CR1⁺ macrophages, AMSC-EVs were added to the cisplatin-stimulated THP-1 cells. As shown in Fig. 5A and B, cisplatin treatment significantly enhanced the expression of pro-inflammatory factors IL-6 and TNF-α, and decreased the expression of anti-inflammatory factor IL-10. Interestingly, AMSC-EVs markedly inhibited the expression of IL-6 and TNF-α, and promoted the expression of the anti-inflammatory factors IL-10 and Arg1. Consistent with the above findings, western blot and immunofluorescence staining analyses further confirmed these results (Fig. 5C and D). Collectively, these data suggest that AMSC-EVs facilitate the conversion of M1 macrophages into M2 macrophages, thereby promoting M2 macrophage-mediated anti-inflammatory responses.

Fig. 5.

AMSC-EVs convert inflammatory macrophages to an M2 phenotype in vitro. A RT-qPCR analysis of TNF-α, IL-6, and IL-10 in cisplatin-treated THP-1 cells (n = 4-5). B RT-qPCR analysis of Arg1 and iNOS in cisplatin-treated THP-1 cells (n = 4-5). C Western blot analysis of proteins associated with macrophage polarization (n = 3). D Immunofluorescence analysis of iNOS and Arg1 expression in cisplatin-treated THP-1 cells (n = 5). Scale bar, 10 μm. Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

The TXNIP/IKKα/NF-κB pathway is involved in AMSC-EVs-mediated CX3CR1⁺ macrophage polarization

In order to detect the molecular mechanism by which AMSC-EVs alleviate AKI damage, we assessed TXNIP protein levels and NF-κB signaling pathway activation. These pathways are essential for macrophage phenotypic polarization and have been previously reported [33, 34]. We found that TXNIP expression was increased in kidney tissues after cisplatin-induced AKI, and AMSC-EVs treatment reversed this effect, as shown by RT-qPCR and Western blot analysis (Fig. 6 A, 6B). Consistent results were obtained using immunofluorescence (Figs. 6E) and flow cytometry (Figure S3A). To further validate our findings, the effect of AMSC-EVs on cisplatin-induced injury was examined in vitro using THP-1 cells. RT-qPCR (Fig. 6 C), Western blot (Fig. 6D), and flow cytometry (Figures S3B) results indicated that TXNIP expression was increased in THP-1 cells after cisplatin treatment, and AMSC-EVs treatment reversed this change.

Fig. 6.

The TXNIP/IKKα/NF-κB pathway is implicated in the polarization of CX3CR1⁺ macrophages induced by AMSC-EVs. A, B Effects of AMSC-EVs on TXNIP expression in kidney tissues of cisplatin-induced AKI mice assessed by RT-qPCR (n = 6) and Western blot (n = 3). C, D Effects of AMSC-EVs on TXNIP expression in THP-1 cell confirmed by RT-qPCR (n = 6) and Western blot (n = 3). E Immunofluorescence staining of CX3CR1⁺TXNIP⁺ macrophages in AKI mice (n = 6), Scale bar, 50 μm. F THP-1 cells transfected with empty vector or pcDNA-TXNIP for 24 hours, TXNIP expression was significantly upregulated in pcDNA-TXNIP-transfected cells (n = 4). G, H RT-qPCR and Western blot analyses of iNOS and Arg1 expression in THP-1 cells following TXNIP overexpression (n = 3). I Western blot analysis of IKKα/NF-κB pathway proteins in kidney tissues of cisplatin-induced AKI mice treated with or without AMSC-EVs (n = 3). J Western blot analysis of IKKα/NF-κB pathway proteins expression in THP-1 cells treated with or without AMSC-EVs (n = 3). K Western blot analysis of IKKα/NF-κB pathway proteins in THP-1 cells after TXNIP overexpression (n = 3). Data are expressed as mean ± SEM, *P < 0.05, **P < 0.01, ***P < 0.001, ****P<0.0001

The thioredoxin (TRX) system is known to play a key role in cellular resistance to oxidative stress [35]. Based on this, we assessed changes in the Trx/Txnrd components. In the AKI model, the expression of Trx1 and Txnrd1 was downregulated, whereas AMSC-EVs treatment restored their expression. Notably, there were no significant differences in the levels of Trx2 and Txnrd2 among the groups (Figures S4A, S4B). This trend was also observed in the THP-1 cell model, consistent with the in vivo results (Figures S4C and S4D).

We then investigated whether AMSC-EVs promote M2 polarization of CX3CR1⁺ macrophages through TXNIP. We established a TXNIP overexpression model in THP-1 cells (Fig. 6F) and found that AMSC-EVs treatment reduced pro-inflammatory gene expression (e.g.,iNOS) while increasing the anti-inflammatory gene Arg1. TXNIP overexpression reversed these effects (Fig. 6G, H). Flow cytometry analysis showed that AMSC-EVs treatment reduced CD86 and increased CD206 expression, promoting M2 polarization, while TXNIP overexpression in the AMSC-EVs + pcDNA-TXNIP group enhanced the M1 phenotype (Supplementary Figure S5A, S5B). These findings suggest that TXNIP is an important target for AMSC-EVs in preventing AKI.

TXNIP has been identified as an important regulator in the signal transduction cascade that leads to NF-κB activation. We tested these major regulators involved in the activation cascade of the NF-κB signaling pathway in vivo (Fig. 6I) and in vitro (Fig. 6J), including IKKα, P-IKKα, p65, and P-p65. P-IKKα and P-p65 expression were significantly increased in the model group and suppressed by AMSC-EVs. Furthermore, TXNIP overexpression substantially attenuated these inhibitory effects (Fig. 6K). All of these results suggest a possible connection between the protective effect of AMSC-EVs and the TXNIP-dependent IKKα/NF-κB signaling pathway.

Discussion

MSCs have emerged as a promising therapeutic candidate for AKI. This study provides in vivo evidence that AMSC-EVs contribute to renal tissue repair and functional restoration after AKI. Notably, our findings imply a potential involvement of CX3CR1⁺ macrophages in the protective mechanisms of AMSC-EVs against cisplatin-induced AKI. We further demonstrated that AMSC-EVs promote the transition of CX3CR1⁺ macrophages from an M1 to an M2 phenotype, consequently attenuating the post-injury inflammatory response in renal tissue. Moreover, AMSC-EVs appear to regulate TXNIP expression, which may suppress M1 polarization while favoring M2 polarization. These effects are potentially mediated through modulation of the TXNIP-IKKα/NF-κB signaling pathway, suggesting a novel mechanism by which AMSC-EVs ameliorate renal inflammation (Graphical Abstract Image).

Macrophages identified in human AKI biopsies correlate with the severity of renal impairment [36, 37]. Singlecell sequencing has revealed the existence of diverse renal macrophage subgroups. In ischemia-reperfusion injury (IRI)-AKI, CD169⁺ macrophages interact with renal vessels, reducing ICAM-1 expression and inhibiting neutrophil accumulation, which helps prevent excessive inflammation [38]. CX3CR1⁺ macrophages are important regulators in kidney injury. Their absence reduces proteinuria and glomerular inflammation [39]. Inhibition of CX3CR1 has been shown to prevent renal fibrosis progression, potentially through M2 polarization and Arg1 upregulation [21]. CX3CR1⁺ macrophages can also promote stem cell proliferation and tissue repair in injury models [40–42]. To explore the role of CX3CR1 ⁺ macrophages in AMSC-EVs therapy, we depleted them using knockout mice. The results indicated that renal macrophage depletion reduced AMSC-EVs efficacy in AKI mice, suggesting that AMSC-EVs may primarily act through CX3CR1⁺ macrophages. CX3CR1⁺ macrophages exhibit both pro-inflammatory and anti-inflammatory effects depending on the tissue, environment, and disease context. In vitro, CX3CL1-CX3CR1 signaling inhibits LPS-induced activation and promotes an anti-inflammatory phenotype [43]. In the cardiovascular system, CX3CR1⁺ macrophages, as resident cardiac macrophages, process abnormal mitochondria via phagocytosis, protecting cardiomyocytes [18]. Studies have shown that AMSC-EVs can promote M2 macrophage polarization and repair kidney damage by modulating macrophages [44]. Our study found that AMSC-EVs downregulate TNF-α, IL-1β, and iNOS while upregulating Arg1 and IL-10 in AKI models. In vivo, AMSC-EVs promote the transformation of CX3CR1⁺ macrophages into an M2 phenotype, enhancing anti-inflammatory repair.

Prior studies have indicated that TXNIP levels are elevated in human proteinuric kidney diseases [23]. Moreover, deletion of TXNIP has been shown to mitigate kidney fibrosis and injury in a murine model of unilateral ureteral obstruction [45]. Significantly, Xiang Li et al. demonstrated that silencing TXNIP substantially reduced the levels of genes associated with the NLRP3/caspase-1 pathway and mitigated kidney injury under septic conditions [46]. Our investigation revealed that TXNIP expression is elevated in mouse renal CX3CR1⁺ macrophages with cisplatin-induced kidney injury, suggesting that TXNIP may contribute to the progression of renal injury. Importantly, TXNIP expression in renal CX3CR1⁺ macrophages was significantly reduced following treatment with AMSC-EVs. In addition, our in vitro experiments demonstrated that macrophages treated with cisplatin exhibited high TXNIP expression, which could be partially reversed by AMSC-EVs. A previous study showed that AMSC-EVs provide protective benefits against acute liver failure in macrophages via miR−17-mediated suppression of TXNIP [27]. Our findings indicate that suppression of TXNIP expression in AKI macrophages can help mitigate renal inflammation and injury. Furthermore, the findings suggest that AMSC-EVs may suppress M1 macrophages and activate M2 macrophages. This implies that AMSC-EVs could reduce TXNIP expression and downregulate the activity of pro-inflammatory M1 macrophages, which is vital for renal protection in AKI.

Mammalian cells possess multiple antioxidant systems to counteract ROS and oxidative injury [47]. The TRX system is an important defense mechanism [35]. Thioredoxin reductase (TrxR) maintains redox balance, while TXNIP binds to Trx1/Trx2, suppressing their reductive activity and enhancing oxidative stress [48, 49]. As an intrinsic inhibitor, TXNIP both limits TRX’s antioxidant role and engages in other signaling pathways. Studies have shown that in IRI-induced liver injury, TRX1 is downregulated while TXNIP and NLRP3 are upregulated [50]; a similar pattern (decreased TRX2 with increased TXNIP/NLRP3) is observed in AngII-induced cardiac remodeling [51]. In various AKI models (IRI, CIN, sepsis), TXNIP is elevated and associated with the mROS-NLRP3 axis [52–54], whereas the TRX system is depleted early in proximal tubules with urinary loss (mainly Trx1/total TRX), accompanied by compensatory enhancement in specific tubular segments or mitochondrial pathways (Trx2/TrxR2) [55–57]. This imbalance of “TXNIP up—TRX down” highlights the potential role of redox dysregulation in AKI pathogenesis. In our cisplatin-induced AKI model, we observed increased TXNIP expression with decreased Trx1/TrxR1, a pattern that was partially reversed by AMSC-EV treatment, suggesting their role in restoring redox balance. Previous studies have shown that ADSC-EVs reduce ROS, activate the Nrf2/HO-1 antioxidant axis, suppress NF-κB-mediated M1 programs, and upregulate SIRT1 to promote M2 polarization, thereby alleviating inflammation and supporting repair [58]. Whether AMSC-EVs confer renal protection through the TXNIP/Trx1/TrxR1 axis and the subsequent promotion of M2 polarization remains an important question for future investigation.

NF-κB activation is a known harmful mechanism in AKI [59, 60]. The IKK complex, composed of catalytic subunits IKKα and IKKβ and regulatory component IKKγ, is essential for NF-κB activation [61]. IKKα/β-mediated phosphorylation and degradation of IκB are crucial for NF-κB nuclear translocation and downstream gene expression. Our prior work showed that IKKα promotes macrophage polarization and fibrotic factor release [62], and its elevated expression exacerbates renal fibrosis via the Wnt/β-catenin pathway in knockout mice [63]. Our current investigation demonstrated that AMSC-EVs inhibit the activation of the IKKα/NF-κB signaling pathway both in vivo and in vitro, suggesting that targeting the IKKα/NF-κB pathway may have potential benefits in managing AKI. Previous studies have highlighted the role of TXNIP in regulating the NF-κB signaling pathway in AKI [24, 64]. Our research showed that overexpression of TXNIP in macrophages significantly activates the IKKα/NF-κB pathway, inhibits M2 macrophage polarization, and enhances M1 macrophage activity in vitro. Consistent with this, overexpression of TXNIP in macrophages reduced the expression of Arg1 and IL-10 while increasing the production and release of IL-6, TNF-α, and iNOS. These findings suggest that AMSC-EVs may reduce TXNIP expression and downregulate pro-inflammatory M1 macrophage activity by modulating the IKKα/NF-κB signaling pathway, which could be important for renal protection in AKI.

This study has several limitations. While we focused on a single AKI model, validation in multiple animal models is necessary to confirm the generalizability of our findings. Additionally, we were unable to use knock-in mice to knock down TXNIP expression in AKI mice to assess renal function, histopathology, and inflammatory markers. The THP-1 cell line also cannot fully replicate the behavior of primary renal CX3CR1⁺ macrophages. Further research is needed to better understand the underlying mechanisms.

In conclusion, our research suggests that AMSC-EVs may modulate the polarization of renal CX3CR1⁺ macrophages and mitigate cisplatin-induced AKI via the TXNIP-IKKα/NF-κB signaling pathway. This study provides new insights into the potential mechanisms of AMSC-EVs-based therapy and highlights its promise as a potential preventive treatment for AKI.

Author contributions

Weijuan Deng performed the majority of experiments and drafted and revised the manuscript. Rui Tang and Menqing Ma performed the biological experiments of western blotting. Hao Zhang assisted the study. Changchun Cao was involved in all aspects of the study conception, design, and direction and provided final approval of the submitted manuscript. All authors reviewed and approved the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (82170698 and 82500834), Youth Science Foundation of Jiangxi Province (20212BAB216071), Health Commission of Jiangxi Province (BKJP1220240788), Jiujiang City Key Research and Development Program Project (2025_000694), and Jiangsu Funding Program for Excellent Postdoctoral Fellows (2025ZB004).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

All animal experiments were conducted in accordance with the ARRIVE 2.0 guidelines and approved by the Animal Ethical and Welfare Committee of Nanjing Medical University (Title: Mechanistic Study of Adipose-derived Mesenchymal Stem Cells Promoting Repair after Acute Kidney Injury; Approval No.: IACUC-2001017; Date: March 4, 2022). This study strictly adhered to the Declaration of Helsinki, and adipose tissue collection was performed after obtaining written informed consent from the participants. The study protocol was approved by the Ethics Committee of Sir Run Run Shaw Hospital, Nanjing Medical University (Title: Protective Effects of Human Adipose-derived Mesenchymal Stem Cell-derived Extracellular Vesicles on Acute Kidney Injury in Mice; Approval No.: 2024-SR-007; Date: January 18, 2024). The THP-1 cell line (RRID: CVCL_0006) was obtained from Shanghai Zhongqiao Xinzou Biotechnology Co., Ltd. (originally purchased from ATCC, USA). For detailed information regarding compliance and prior ethical approval of the cell line, please visit the ATCC official website: https://www.atcc.org.

Competing interests

The authors declare no competing interests.