Extracellular vesicle-derived TMEM106A participated in podocyte injury via EGR1 in preeclampsia

Abstract

Preeclampsia (PE), a pregnancy-specific hypertensive disorder affecting approximately 5–8% of pregnancies, is a leading cause of maternal and fetal morbidity and mortality. Although renal complications, particularly podocyte injury, are hallmarks of preeclampsia-related organ dysfunction, the underlying molecular mechanisms remain poorly understood. In this study, PE patients from the Gene Expression Omnibus (GEO) database (GSE192902) and injured podocytes from GSE124622 were used to select TMEM106A. Quantitative PCR (qPCR) revealed that TMEM106A expression was significantly downregulated in the urine of PE patients, and urinary TMEM106A levels were negatively correlated with nephrin, PCX, IL-1β, and TNF-α levels. Compared with the controls, podocytes exposed to extracellular vesicles (EVs) from PE patients exhibit reduced TMEM106A levels and aggravated podocyte injury. Silencing TMEM106A in vitro aggravated podocyte cytoskeleton rearrangement and inflammation, whereas its overexpression alleviated these responses. Transcriptomic analysis identified EGR1 as a downstream gene of TMEM106A. Mechanistically, TMEM106A silencing promoted podocyte injury via EGR1 upregulation. In vivo, podocyte-specific overexpression of TMEM106A alleviated renal injury in an L-NAME-induced PE mouse model, whereas podocyte-specific deletion of EGR1 conferred renal protection in the same model. These findings demonstrate that EV-derived TMEM106A modulates podocyte injury in PE by regulating EGR1 expression, highlighting a potential therapeutic axis for mitigating renal dysfunction in preeclampsia.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00018-026-06272-4.

Introduction

Preeclampsia (PE) is a multisystem disorder of pregnancy that affects approximately 3–8% of pregnant women worldwide and remains a leading cause of maternal and fetal morbidity and mortality [1–4]. Clinically, PE is characterized by new-onset hypertension after 20 weeks of gestation, accompanied by proteinuria or other signs of end-organ dysfunction [5–7]. Accumulating evidence indicates that podocyte injury or detachment plays a key role in the development of proteinuria in patients with PE [8, 9]. However, the specific mechanism leading to podocyte injury remains largely unclear.

In recent years, extracellular vesicles (EVs) have attracted attention as important mediators of intercellular communication [10–14]. In PE, placental-derived EVs are markedly elevated in the maternal circulation and contain many cellular components, such as DNA, RNA, lipids, metabolites, and cytosolic and cell-surface proteins, which are capable of modulating the function of target cells [15–19]. Previous studies have shown that EVs from the serum of PE patients can be internalized by glomerular endothelial cells, leading to impaired proliferation, migration, and tube formation, ultimately compromising endothelial barrier integrity and renal function [16, 20–22]. Although previous studies have highlighted the potential role of circulating EVs in PE, it remains unclear whether they directly mediate podocyte injury [23–25].

To explore potential mediators of podocyte injury in PE, we performed differential gene expression analysis on two independent transcriptomic datasets, one derived from PE patient plasma (GSE192902) [26] and the other from PAN-injured and Adriamycin-injured human podocytes (GSE124622) [27, 28]. Six overlapping genes were identified, and TMEM106A was selected for subsequent functional analysis. TMEM106A is a type II transmembrane glycoprotein that has previously been described as a tumor suppressor with immunomodulatory functions [29–31]. Prior studies have reported that the TMEM106 family member TMEM106B is localized to endolysosomal and late lysosomal compartments, where it plays a role in regulating endolysosomal dynamics and vesicular trafficking [32]. Emerging evidence indicates that this family can mediate intercellular communication through extracellular vesicle (EV) transport. On the basis of previous research, we aimed to explore whether TMEM106A is involved in the pathophysiological mechanism of PE in patients through the use of plasma-derived EVs.

In this study, we aimed to investigate the role of TMEM106A in podocyte injury and proteinuria in PE. We examined TMEM106A expression in plasma and urine samples from PE patients and assessed its functional effects on cultured podocytes. Using a PE mouse model, we further evaluated the in vivo significance of TMEM106A overexpression. Our findings demonstrate that the aberrant expression of EV-derived TMEM106A directly induces podocyte injury in PE, suggesting that targeting TMEM106A may be an innovative therapeutic strategy for the prevention and treatment of podocyte injury in PE in the future.

Materials and methods

Patient samples

The blood and urine samples used in the study were all from the Department of Obstetrics and Gynecology of the Second Hospital of Shandong University. The study included two groups, namely, a healthy pregnancy group (control) and a PE group, which were diagnosed according to established criteria. All participants provided written informed consent prior to enrollment. The study was approved by the Ethics Committee of The Second Hospital of Shandong University.

Urine samples were collected for analysis of TMEM106A expression, podocyte-specific proteins, and inflammatory cytokines. Plasma samples were used for extracellular vesicle isolation and systemic inflammatory marker analysis. The participants in the control group were matched for maternal age and gestational age.

Cell culture and treatment

Human podocyte culture

Conditionally immortalized human podocytes (HPCs) were kindly provided by Dr. Peter Mundel [33]. Podocytes were cultured in RPMI 1640 medium (Gibco, USA) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin at 37 °C in a humidified incubator with 5% CO2.

Transfection of human podocytes

For gene overexpression and silencing studies, podocytes were transfected using Lipofectamine 3000 (Invitrogen, USA) following the manufacturer’s instructions. The siRNAs we used were all purchased from GenePharma: siRNA targeting TMEM106A (si-TMEM106A), siRNA targeting EGR1 (si-EGR1) and their corresponding control siRNA (si-NC). Additionally, the EGR1 overexpression plasmid was used, which was purchased from Genomeditech (Shanghai, China) Transfections were performed at 70–80% confluency, and the cells were incubated with the transfection complexes for 48 h. The sequences are as follows:

TMEM106A siRNA 5’-GCAGUUGGUGGCUCUCAUUTT-3’;

EGR1 siRNA 5’-AGUUUGCCAGGAGCGAUGA-3’;

NC siRNA, 5’-UUCUCCGAACGUGUCACGUTT-3’.

Plasma EV Isolation and trafficking assay

Isolation and characterization of EVs

EVs were isolated from plasma samples obtained from PE patients and gestational age-matched healthy controls. Briefly, plasma was separated from whole blood via sequential centrifugation (1500×g, 15 min; 12,000×g, 30 min; 4 °C) and diluted 1:1 with phosphate-buffered saline (PBS, pH 7.4). The EVs were pelleted by ultracentrifugation at 100,000 × g for 70 min at 4 °C (Himac cp100wx, HITACHI, Japan), washed in PBS under identical conditions, and resuspended in 100 µL of PBS for storage at − 80 °C. EV protein identity was confirmed by western blot analysis with antibodies targeting CD63 and calnexin.

EV trafficking assay

To investigate EV internalization dynamics, isolated EVs were fluorescently labeled with the lipophilic tracer PKH26 (Sigma‒Aldrich, USA) per the manufacturer’s protocol. Excess dye was quenched by incubating the labeled EV suspension with 1% bovine serum albumin (BSA) for 10 min, followed by purification through ultracentrifugation (120,000 × g, 2 h, 4 °C). The purified PKH26-EV pellet was resuspended in phosphate-buffered saline (PBS) and seeded onto recipient cells for 24–48 h. Following coculture, the cells were fixed with 4% paraformaldehyde (PFA), permeabilized with 0.1% Triton X-100, and costained with Actin-Tracker Green (Beyotime, China) and DAPI to visualize the cytoskeletal architecture and nuclear morphology. Intracellular EV localization was verified using a laser-scanning confocal microscope (Olympus, Japan).

Gene expression and protein analysis

RNA extraction and qPCR

Total RNA was extracted from podocytes using TRIzol reagent (Takara, Japan). For plasma RNA, a plasma RNA extraction kit (BIOG Biotechnology, China) was used; for urine samples, a Urine RNA extraction kit (BIOG Biotechnology, China) was used. RNA purity and concentration were assessed using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, USA).

Complementary DNA was synthesized using the PrimeScript™ RT Reagent Kit with gDNA Eraser (Takara, Japan). qPCR was conducted using TB Green^® Premix Ex Taq™ II (Takara) on a LightCycler^® 480 System (Roche, Switzerland). β-actin was used as an internal control. The primer sequences are listed in Supplementary Table S1.

Western blotting

Total protein was extracted from podocytes, plasma EVs, and glomerulus using RIPA lysis buffer (Beyotime, China) supplemented with protease inhibitors. The protein concentration was determined by a BCA assay. Equal amounts (30 µg) were resolved by SDS-PAGE and transferred onto PVDF membranes (Millipore, USA). The membranes were blocked in 5% nonfat milk and incubated with primary antibodies overnight at 4 °C, followed by incubation with HRP-conjugated secondary antibodies for 1 h at room temperature. Protein bands were detected using enhanced chemiluminescence (ECL) reagent (Millipore) and imaged using an Amersham Imager 680 (GE Healthcare). β-actin served as a loading control.

The following antibodies were used: CD63 (Abcam, ab134045), calnexin (Proteintech, 10427-2-AP), TMEM106A (NOVUS, NBP2-41165), Synaptopodin (Proteintech, 21064-1-AP), ZO-1 (Proteintech, 21773-1-AP), podocin (Absin, abs126305), IL-1β (Proteintech, 16806-1-AP), IL-6 (Proteintech, 66146-1-Ig), TNF-α (Proteintech, 60291-1-Ig), EGR1(Diagbio, db15021) and β-actin (Proteintech, 66009-1-Ig).

Enzyme-linked immunosorbent assay (ELISA)

Urine samples were collected and centrifuged at 3000 × g for 10 min at 4 °C to remove insoluble debris. The supernatants were aliquoted and stored at − 80 °C until analysis. Levels of IL-6, IL-1β, TNF-α, nephrin, and podocalyxin in urine were measured using commercial high-sensitivity human ELISA kits (IL-6: JONLNBIO, JL14113; IL-1β: Elabscience, E-EL-H0149; TNF-α: Elabscience, E-EL-H0109; nephrin: JONLNBIO, JL15363; podocalyxin: JONLNBIO, JL20100) according to the manufacturer’s instructions. Additionally, IL-6 levels in podocytes were also measured using the same ELISA kit (JONLNBIO, JL14113). The absorbance was measured at 450 nm with a reference wavelength of 570 nm using a microplate reader (Thermo Scientific Multiskan FC).

Cellular immunofluorescence staining

Podocytes were fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, and stained with primary antibodies against synaptopodin, ZO-1, and podocin. The staining was visualized using a laser-scanning confocal microscope (FV3000; Olympus, Japan). The antibodies used were Synaptopodin (Proteintech, 67339-1-Ig), ZO-1 (Proteintech, 21773-1-AP), and Podocin (Absin, abs126305).

L-NAME-Induced PE mouse model

Animal model and grouping

Pregnant C57BL/6 mice were used. PE was induced by daily intraperitoneal injections of L-NAME (100 mg/kg/day; Sigma-Aldrich, USA) or PBS from E10.5 to E16.5.

To achieve TMEM106A overexpression in podocytes, recombinant adeno-associated virus serotype 9 (AAV9) (WZ Biosciences, Shanghai, China) encoding TMEM106A under the podocyte-specific Nphs1 promoter (AAV9-TMEM106A) or a control virus (AAV9-NC) was administered via tail vein injection at E10.5. For the EGR1 knockout model, EGR1^flox/flox mice were crossed with Nphs2-Cre transgenic mice to generate podocyte-specific EGR1 knockout (EGR1^KO) mice, which, together with their wild-type (WT) littermates, were also subjected to the PE induction protocol described above.

Sample collection and analysis

At E17.5, pregnant mice were euthanized, and urine, kidney, placenta, and fetus samples were collected for analysis. Blood pressure measurements were taken using the tail cuff method as previously described [34]. Immunofluorescence staining was performed on kidney sections to evaluate synaptopodin, ZO-1, F4/80, TMEM106A, and EGR1. Western blotting was performed on glomeruli to assess IL-1β, TNF-α, IL-6, TMEM106A and EGR1 expression. Mouse primary podocytes were isolated from glomeruli outgrowths as described previously [35]. Urine protein levels were measured to evaluate renal injury. All data were quantified using ImageJ software.

Statistical analysis

All experiments were repeated independently at least three times. The data are presented as the means ± standard deviations (SDs). Statistical analysis was performed using GraphPad Prism 9.0 (GraphPad Software, USA). Intergroup comparisons were made using unpaired Student’s t tests, one-way ANOVA or two-way ANOVA, followed by Tukey’s multiple comparison test. Correlation analysis was performed using Pearson’s correlation coefficient. A two-sided P value < 0.05 was considered to indicate statistical significance.

Results

Transcriptional expression of TMEM106A is downregulated in the plasma of PE patients and in podocyte injury models

To identify genes involved in podocyte injury in PE, we analyzed transcriptomic data from two GEO datasets. For PE, RNA expression profiles were obtained from GSE192902, which included 49 PE samples and 238 healthy controls. Differential expression analysis (|log2FC| ≥ 1, p < 0.05) revealed 98 upregulated and 523 downregulated genes in the PE group (Fig. 1A).

Fig. 1.

Transcriptomic analysis reveals gene expression in preeclampsia and podocyte injury models. A Differentially expressed genes (DEGs) in cell-free RNA from preeclampsia (PE) patients compared to healthy controls (GSE192902). Left: Volcano plot showing 98 significantly upregulated (red) and 523 downregulated (blue) genes (|log₂FC| ≥ 1, P < 0.05). Right: Heatmap of representative DEGs; the color scale indicates z-scores of log₂(normalized counts + 1). B Transcriptomic changes in podocytes treated with Adriamycin (GSE124622). Left: Volcano plot displaying 1973 upregulated and 1037 downregulated DEGs (|log₂FC| ≥ 1, padj < 0.05). Right: Corresponding heatmap. C Transcriptomic changes in podocytes treated with PAN (GSE124622). Analysis using the same parameters as in (B) identified 5509 upregulated and 3057 downregulated DEGs. The associated heatmap is shown on the right. D Venn diagram illustrating the overlap of DEGs from Adriamycin-treated and PAN-treated podocytes. The intersection contains 2007 shared genes, defined as the core podocyte injury-associated gene set (Set B). E Venn diagram showing the intersection between PE-associated DEGs (Set A) and the podocyte injury gene set (Set B, derived from D). This overlap identifies 69 genes commonly dysregulated in both PE and podocyte injury

Furthermore, we used GSE124622, which includes gene expression data from human podocytes treated with either Adriamycin or PAN. Compared with the control podocytes, 1973 genes were upregulated and 1037 genes were downregulated in the Adriamycin-induced podocyte injury model (|log2FC| ≥ 1, padj < 0.05). In the PAN-induced kidney injury model, 5509 genes were upregulated, and 3057 genes were downregulated (|log2FC| ≥ 1, padj < 0.05) (Fig. 1B–C). The intersection of DEGs from the Adriamycin and PAN treatments defined a podocyte injury-associated gene set (B) containing 2007 overlapping genes (Fig. 1D).

We next intersected the PE DEGs (A) with the podocyte injury gene set (B), identifying 69 genes that are jointly associated with PE and podocyte injury (Fig. 1E). To increase biological relevance, we further filtered for genes whose expression trends were consistent between the two datasets. This yielded 6 consistently downregulated genes, including CLTCL1, COL1A1, SLC37A3, WDFY3, CLMN and TMEM106A. Among them, TMEM106A was selected for further validation because of its potential role in renal pathology.

TMEM106A is downregulated in the urine of preeclamptic patients and is correlated with renal injury and inflammation

To validate the clinical relevance of TMEM106A in PE, urine samples and clinical data were prospectively collected from a cohort of 20 PE patients and 20 gestational age-matched healthy controls (Normal). Baseline maternal demographics, including age, BMI, and gestational age at sampling, were not significantly different across the groups (all P values > 0.05). The PE cohort exhibited the following characteristic disease manifestations: elevated systolic/diastolic blood pressure, increased serum creatinine, a reduced estimated glomerular filtration rate (eGFR, P < 0.05), and elevated 24-hour urinary protein excretion (P < 0.05), confirming renal dysfunction (Supplementary Table S2). qPCR analysis of urinary mRNA revealed significant downregulation of TMEM106A in PE patients compared with controls (Fig. 2A). Consistent with the results of podocyte injury, the results of an ELISA revealed that urinary concentrations of the podocyte-specific structural markers podocalyxin (PCX) and nephrin were significantly increased in the PE group (Fig. 2B). Given the recognized role of inflammation in podocyte injury, we profiled inflammatory cytokines in urine. qPCR analysis revealed significantly elevated mRNA levels of interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α) in PE patients (Fig. 2C). Similarly, ELISA confirmed increased urinary protein concentrations of IL-1β, TNF-α, and interleukin-6 (IL-6) (Fig. 2D).

Fig. 2.

TMEM106A is downregulated in the urine of preeclamptic patients and is correlated with renal injury and inflammation. A qPCR analysis of TMEM106A mRNA levels in urine samples from healthy controls (Normal, n=20) and patients with preeclampsia (PE, n=20). B Urinary levels of podocyte injury markers measured by ELISA. C Analysis of inflammatory cytokine mRNA expression. The relative mRNA level of urinary IL-1β and TNF-α were determined by qPCR. D Analysis of inflammatory cytokine protein levels. The concentration of urinary IL-1β, TNF-α, and IL-6 protein were measured by ELISA. E Nephrin (r = -0.4904, P = 0.0282) and podocalyxin (PCX, r = -0.4769, P = 0.0335), and significant negative correlations with the inflammatory markers IL-1β (r = -0.5213, P = 0.0184) and TNF-α (r = -0.5652, P = 0.0094).Data are presented as mean ± SD (n=20 subjects per group); *P < 0.05, **P < 0.01, ***P < 0.001 by Student’s t-test; correlation analyzed by Pearson’s correlation

Pearson correlation analysis further revealed significant associations between urinary TMEM106A levels and key biomarkers. TMEM106A expression was negatively correlated with the expression of the podocyte injury markers nephrin (r = -0.4904; P = 0.0282) and PCX (r = -0.4769; P = 0.0335), as well as with the inflammatory cytokines IL-1β (r = -0.5213; P = 0.0184) and TNF-α (r = -0.5652; P = 0.0094) (Fig. 2E). Collectively, these findings indicate that the downregulation of TMEM106A is associated with podocyte injury and inflammatory dysregulation in PE.

EVs from PE patients exhibit reduced TMEM106A expression and promote podocyte injury

Previous studies have shown that EVs play a role in the renal function of patients with PE, and plasma-derived EVs were isolated from 20 PE patients and 20 gestational age-matched healthy controls. EV purity was validated by western blotting, confirming enrichment of the canonical surface marker CD63 (EV-specific) and the absence of calnexin (an endoplasmic reticulum marker), thereby excluding cellular contamination (Fig. 3A). In addition, western blot analysis demonstrated significantly lower TMEM106A protein levels in the PE EVs than in the control EVs (Fig. 3B).

Fig. 3.

EVs from PE patients exhibit reduced TMEM106A expression and promote podocyte injury. A Western blot analysis of exosomal markers (CD63) and the negative maker calnexin in extracellular vesicles (EVs) isolated from plasm of healthy controls (Normal^EV) and preeclampsia patients (PE^EV). B Western blot and quantitative analysis of TMEM106A protein levels in EVs from Normal^EV and PEEV. C (Left) Representative fluorescence microscopy images of podocytes at indicated time points after incubation with PKH26-labeled EVs (red) isolated from Normal^EV or PE^EV. Cytoskeletal actin and nuclei were stained with Actin Tracker Green (green) and DAPI (blue). Merged images are shown in the rightmost column. Scale bar, 50 μm. (Right) Quantification of internalized EVs. The relative fluorescence intensity of internalized PKH26 was measured to assess EV uptake over time. D Quantitative analysis of TMEM106A protein levels in podocytes. E Immunofluorescence analysis of TMEM106A in podocytes. Representative fluorescence images of podocytes treated with EVs from NormalEV or PEEV. Cells were stained for TMEM106A (green) and nuclei (DAPI, blue). Scale bar, 50 μm. F Western blot and quantitative analysis of Synaptopodin and Podocin in podocytes treated with Normal^EV or PE^EV. G PE-EVs upregulate inflammatory cytokines. Western blot and quantitative analysis of IL-1β, IL-6, and TNF-α in podocytes treated with NormalEVEV or PEEVEV. Data are presented as mean ± SD from 3 independent biological replicates; β-actin served as the loading control. *P < 0.05, **P < 0.01, ***P < 0.001 by Student’s t-test for panels B, C, and D; two-way ANOVA for panels F and G

Furthermore, EVs derived from PE patients were fluorescently labeled with PKH26 and incubated with primary human podocytes. As shown in Fig. 3C, cytoplasmic internalization of labeled EVs was detectable in podocytes within 24 h, with the fluorescence intensity significantly increasing by 48 h, revealing that human podocytes exhibit robust internalization capacity for PE-derived EVs.

Strikingly, exposure to PE^EV, but not Normal^EV, led to a significant downregulation of TMEM106A protein in recipient podocytes, as confirmed by both western blotting (Fig. 3D) and immunofluorescence staining (Fig. 3E). These findings indicate that TMEM106A-deficient EVs can suppress the expression of TMEM106A in target cells.

We then evaluated the functional consequences of this EV-mediated downregulation of TMEM106A expression. Treatment with PE^EV significantly reduced the protein levels of the podocyte-specific structural proteins synaptopodin and podocin (Fig. 3F), indicating disruption of podocyte structural integrity. Concurrently, PE^EV treatment markedly upregulated the expression of the inflammatory cytokines IL-1β, IL-6, and TNF-α at the protein level (Fig. 3G). Collectively, these data demonstrate that TMEM106A-deficient EVs from PE patients are efficiently internalized by podocytes, where they downregulate TMEM106A expression and subsequently drive podocyte injury.

TMEM106A deficiency induces podocyte injury

To elucidate the functional role of TMEM106A in maintaining podocyte homeostasis, we performed targeted knockdown of TMEM106A in cultured human podocytes using small interfering RNA. Cells were transfected with either TMEM106A-specific siRNA or a nontargeting negative control. The efficiency of gene silencing was confirmed at both the transcriptional and protein levels by qPCR (Fig. 4A) and western blot analysis (Fig. 4B), respectively, which demonstrated a robust reduction in TMEM106A expression in the si-TMEM106A group.

Fig. 4.

TMEM106A Deficiency Induces Podocyte Injury. A qPCR analysis of control (con) podocytes, podocytes transfected with TMEM106A-specific siRNA (si-TMEM106A) or non-targeting control siRNA (si-NC). B Representative Western blot of TMEM106A (Upper) and quantitation (Lower) in con, si-NC and si-TMEM106A podocytes. C Relative mRNA levels of the inflammatory cytokines IL-1β, IL-6, and TNF-α, measured by qPCR. D Western blot and quantitative analysis of the inflammatory cytokine proteins IL-1β, IL-6, and TNF-α. E Relative mRNA level of the critical podocyte-specific structural proteins synaptopodin and ZO-1. F Western blot and quantitative analysis of the proteins ZO-1, Synaptopodin, and Podocin. G Immunofluorescence staining of synaptopodin (green, left), ZO-1 (red, middle), and podocin (red, right) in podocytes. Nuclei were stained with DAPI (blue). H Quantitative analysis of the cortical F-actin scores (CFS) among the si-TMEM106A, con and si-NC groups. The quantification of CFS was performed using ImageJ software. I ELISA and Western blot based quantification of IL-6, TNF-a and IL1β in protein lysates of conditioned medium. Data are presented as mean ± SD from 3 independent biological replicates; β-actin served as the loading control. Statistical significance was determined by one-way ANOVA for the panels A, B, C, E, H, and the left panel of I, and by two-way ANOVA for the panels D, F, and the right panel of I. *P < 0.05, **P < 0.01, ***P < 0.001

Functionally, TMEM106A knockdown significantly upregulated the mRNA expression of the inflammatory cytokines IL-1β, IL-6, and TNF-α (Fig. 4C). Consistent with this, the protein levels of IL-1β, IL-6, and TNF-α were also substantially increased (Fig. 4D). Concurrently, at the transcriptional level, qPCR analysis revealed significant downregulation of synaptopodin and ZO-1 mRNA expression (Fig. 4E). These results were further corroborated at the protein level, where western blot analysis revealed a concomitant decrease in critical podocyte-specific structural proteins, namely, ZO-1, synaptopodin, and podocin, which are essential for maintaining the integrity of the glomerular filtration barrier (Fig. 4F). Immunofluorescence staining visually revealed disruptions in the expression of synaptopodin, ZO-1, and podocin upon knockdown (Fig. 4G).

Given the established link between inflammation and cytoskeletal integrity, we performed immunofluorescence staining using phalloidin to examine the cytoskeleton, revealing marked disorganization and loss of filamentous structure in TMEM106A-deficient cells (Fig. 4H), reinforcing the structural damage observed in the absence of TMEM106A. These findings indicate that TMEM106A plays a central role in preserving podocyte structural stability and in suppressing inflammatory responses under pathological conditions. Its loss may contribute to the glomerular injury observed in PE, potentially through its effects on inflammatory signaling and the structural integrity of podocytes.

Analysis of the conditioned medium revealed that TMEM106A knockdown significantly increased the secretion of the IL-6 protein, as measured by ELISA. Furthermore, western blot analysis of acetone-precipitated proteins from the conditioned medium confirmed elevated levels of secreted TNF-α and IL-1β (Fig. 4I). Together, these data highlight the importance of TMEM106A in maintaining podocyte integrity and preventing inflammatory responses that contribute to renal injury.

EGR1 mediates TMEM106A deficiency-induced podocyte injury

To further investigate the molecular mechanisms through which TMEM106A regulates podocyte function, we conducted transcriptomic profiling using RNA sequencing in si-NC- and si-TMEM106A-transfected podocytes. Differential expression analysis revealed 317 significantly altered genes, including 74 upregulated and 243 downregulated genes (fold change ≥ 2, padj < 0.05) (Fig. 5A). Among the most prominently upregulated genes, EGR1, a transcription factor implicated in stress responses, inflammation, and glomerular injury, emerged as a key candidate (Fig. 5B). Consistent with this observation, ChEA3 transcription factor enrichment analysis identified EGR1 as a key regulator, predicted to control approximately 13.6% of the differentially expressed genes (P = 0.001768; top 10 ranked TFs are listed in Supplementary Table S3), indicating its potential involvement in podocyte injury. Importantly, these predicted regulatory roles of EGR1 are consistent with the experimental evidence obtained in our study.

Fig. 5.

EGR1 Mediates TMEM106A Deficiency-Induced Podocyte Injury. A Volcano plot of differentially expressed genes (DEGs) in TMEM106A knockdown (si-TMEM106A) and control (si-NC) podocytes. The plot shows log2 fold change (x-axis) versus -log10 padj (y-axis) for each gene. Upregulated genes are shown in red (74 genes), downregulated genes in blue (243 genes), and nonsignificant genes in black (12295 genes). The horizontal dashed line represents the padj significance threshold and the vertical dashed lines represent significance threshold for log2FC. B Heatmap of the most significantly altered genes (based on fold change) in TMEM106A knockdown podocytes. Color scale indicates gene expression levels (red = upregulated, blue = downregulated). C qPCR assessment of relative mRNA levels of EGR1. D Western blot and quantitative analysis showing the protein levels of EGR1 in con and si-TMEM106A group. E Immunofluorescence staining of EGR1 (red) in podocytes. Nuclei were stained with DAPI (blue). Scale bar, 50 µm. F Violin plot of EGR1 expression in between control and injured podocytes from GSE124622. G Relative mRNA levels of the inflammatory cytokines IL-1β, IL-6, and TNF-α, measured by qPCR. H Western blot and quantitative analysis of the inflammatory cytokine proteins IL-1β, IL-6, and TNF-α. I Relative mRNA level of the critical podocyte-specific structural proteins synaptopodin and ZO-1 in cultured podocytes overexpressing EGR1 (OE-EGR1) compared to control (OE-NC). J Western blot and quantitative analysis of the proteins ZO-1, Synaptopodin, and Podocin. K Immunofluorescence staining of synaptopodin. ​Representative images showing the expression and localization of the podocyte slit diaphragm protein synaptopodin (green) in podocytes. Nuclei were stained with DAPI (blue). Merged images are shown on the right. Scale bar, 50 µm. L ​(Left) Representative immunofluorescence images of filamentous actin (F-actin, green) in podocytes. Nuclei were stained with DAPI (blue). (Right) Quantitative analysis of the mean cortical F-actin score (Mean CFS). Scale bar, 50 µm. Quantification was performed using ImageJ software. M Western blot analysis showing the expression of podocyte-specific markers synaptopodin and podocin, and inflammatory cytokines IL-6 and TNF-α, in con and si-TMEM106A with or without EGR1 co-silencing. Data are presented as mean ± SD from 3 independent biological replicates; β-actin served as the loading control. Statistical significance was determined by Student’s t-test for panels C and D, by one-way ANOVA for panels G, I, and L, and by two-way ANOVA for panels H and M. *P < 0.05, **P < 0.01, ***P < 0.001

We validated these findings by qPCR and western blotting and confirmed that TMEM106A knockdown led to significant increases in both EGR1 mRNA and EGR1 protein levels (Fig. 5C–D). Immunofluorescence staining further demonstrated enhanced nuclear localization of EGR1 in TMEM106A knockdown podocytes (Fig. 5E), supporting its potential role in mediating the effects of TMEM106A loss on podocyte function. Notably, analysis of a previously identified TMEM106A-downregulated podocyte injury dataset (GSE124622) showed that EGR1 expression was significantly increased in PAN-induced podocyte injury model compared to controls (Fig. 5F). Consistently, we analyzed another dataset comprising a rat model of PE (GSE173380) and found that TMEM106A trended towards downregulation while EGR1 expression increased significantly (data not shown), supporting a role a connection for TMEM106A-EGR1 axis in PE.

To assess the functional role of EGR1 in mediating the effects of TMEM106A, we overexpressed EGR1 in cultured podocytes. EGR1 overexpression recapitulated the inflammatory and structural phenotypes observed in TMEM106A knockdown podocytes. Specifically, the levels of IL-1β, IL-6, and TNF-α were significantly elevated at both the mRNA and protein levels (Fig. 5G-H), whereas the expression of podocyte-specific structural proteins, such as synaptopodin, podocin, and ZO-1, was notably reduced (Fig. 5I-J). Immunofluorescence staining confirmed the marked reduction in synaptopodin expression (Fig. 5K), and quantification of the cortical F-actin score (CFS) revealed significant cytoskeletal disorganization upon EGR1 overexpression (Fig. 5L). These findings suggest that EGR1 plays a pivotal role in driving the inflammatory response and disrupting podocyte integrity.

To further investigate whether EGR1 functions as a downstream effector of TMEM106A, we conducted rescue experiments by silencing EGR1 in TMEM106A knockdown podocytes (Fig. 5M). This intervention significantly suppressed the upregulation of IL-6 and TNF-α expression and partially restored the expression of synaptopodin and podocin, suggesting that EGR1 mediates the inflammatory and structural defects induced by TMEM106A knockdown.

In summary, these findings strongly support the hypothesis that TMEM106A exerts its protective effects in podocytes, at least in part, by restraining the EGR1-mediated activation of inflammatory and injury pathways. Thus, EGR1 serves as a critical downstream effector in the TMEM106A signaling pathway, influencing podocyte integrity and inflammatory responses in glomerular injury.

TMEM106A overexpression alleviates podocyte and glomerular injury in L-NAME-induced PE mice

To evaluate the therapeutic potential of TMEM106A in vivo, we employed an L-NAME-induced mouse model of PE. Pregnant mice were administered L-NAME from embryonic day (E) 10.5 to E16.5 and received a single injection of adeno-associated virus 9 (AAV9) carrying TMEM106A under the podocyte-specific Nphs1 promoter (AAV9-TMEM106A) or a control vector (AAV9-NC) at E10.5 to induce podocyte-specific overexpression (Fig. 6A). The overexpression efficiency is shown in Supplementary Figure S1A. Immunofluorescence staining confirmed decreased expression of TMEM106A and its colocalization with the podocyte marker synaptopodin in glomeruli (Fig. 6B, Supplementary Figure S1B).

Fig. 6.

TMEM106A Overexpression Alleviates Podocyte and Glomerular Injury in L-NAME-Induced PE Mice. A Schematic timeline showing the experimental procedure for inducing preeclampsia (PE) in C57BL/6 pregnant mice. Mice were administered either PBS or L-NAME (100 mg/kg/day) intraperitoneally from embryonic day (E) 10.5 to E16.5. AAV9 vectors encoding either TMEM106A (AAV9-TMEM106A) or control vectors (AAV9-NC) were injected at E10.5 to induce overexpression of TMEM106A. B Representative immunofluorescence images of kidney sections stained for the podocyte marker synaptopodin (green) and TMEM106A (red). Nuclei were stained with DAPI (blue). Merged and enlarged views show the co-localization and expression levels. Scale bar, 50 µm. C Systolic blood pressure measurements throughout the experimental timeline. (n=6 per group). D BCA protein quantification was used to assess urinary protein levels in control and L-NAME-treated mice with or without TMEM106A overexpression. L-NAME treatment significantly increased urinary protein excretion. AAV9-TMEM106A significantly reduced urinary protein levels. E Fetal and placental weights from L-NAME-treated mice with or without TMEM106A overexpression. F Hematoxylin and eosin (HE) staining of kidney sections from control and L-NAME-treated mice with or without TMEM106A overexpression. G Transmission electron microscopy images showing podocyte foot process width and glomerular basement membrane (GBM) thickness in L-NAME-treated mice with or without TMEM106A overexpression. H Immunofluorescence staining for podocyte-specific markers synaptopodin (green) and ZO-1 (red) in glomerulus from control and L-NAME-treated mice. I Western blot analysis showing the expression of inflammatory cytokines TNF-α and IL-1β in glomerulus from L-NAME-treated mice with or without TMEM106A overexpression. Additionally, EGR1 expression was also assessed, showing that TMEM106A overexpression led to a reduction in EGR1 levels in L-NAME-treated mice. J Immunofluorescence staining for F4/80 in kidney sections from control and L-NAME-treated mice. Data are presented as mean ± SD from 6 independent biological replicates; β-actin served as the loading control. Statistical significance was determined by two-way ANOVA followed by Tukey’s multiple comparison test. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001 versus the Con + AAV9-NC group; ^#P <0.05, ^##P <0.01, ^###P <0.001 versus the L-NAME + AAV9-NC group

Notably, TMEM106A overexpression significantly attenuated the PE phenotype. Compared to L-NAME mice administered AAV9-NC, those injected with AAV9-TMEM106A exhibited a decreased systolic blood pressure (Fig. 6C) and urinary protein excretion (Fig. 6D). Additionally, fetal and placental weights were notably increased in the AAV9-TMEM106A group compared to the AAV9-NC group within the L-NAME-treated mice (Fig. 6E), suggesting that TMEM106A may contribute to placental function and fetal development. Histological analyses revealed preserved glomerular architecture in PE mice injected with AAV9-TMEM106A, as evidenced by HE staining (Fig. 6F) and electron microscopy, which revealed reduced podocyte foot process effacement and decreased glomerular basement membrane (GBM) thickness (Fig. 6G). At the molecular level, immunofluorescence staining revealed that the expression of the podocyte-specific structural proteins synaptopodin and ZO-1 was restored in PE mice injected with AAV9-TMEM106A (Fig. 6H). These findings indicate that TMEM106A overexpression contributes to the preservation of podocyte integrity under PE conditions.

Moreover, compared to PE mice injected with AAV9-NC, those injected with AAV9-TMEM106A decreased expression of the inflammatory cytokines IL-1β and TNF-α (Fig. 6I, Supplementary Figure S1C) and led to a significant reduction in renal infiltration of F4/80-positive macrophages (Fig. 6J, Supplementary Figure S1D). Western blot analysis, as shown in Fig. 6I, also revealed increased expression of EGR1 in L-NAME-treated mice, which was significantly reduced by TMEM106A overexpression, indicating that TMEM106A may exert its renoprotective effects through modulation of the EGR1 pathway. Collectively, these data indicate that TMEM106A has renoprotective effects on PE by mitigating glomerular inflammation, preserving podocyte function, and improving overall kidney structure.

Podocyte-specific deletion of EGR1 protects against podocyte and glomerular injury in PE mice

To further elucidate the pathological role of EGR1 in podocyte injury during PE, we generated podocyte-specific EGR1^KO mice by crossing floxed EGR1 mice with Nphs2-Cre transgenic mice. Both EGR1^KO and WT littermate controls were subjected to L-NAME-induced PE from E10.5 to E16.5 (Fig. 7A). In WT mice, L-NAME treatment robustly upregulated EGR1 expression in glomeruli, as shown by immunofluorescence costaining for the podocyte marker synaptopodin (Fig. 7B). Western blot analysis of primary podocytes confirmed the successful deletion of EGR1 specifically in the podocytes of EGR1^KO mice (Fig. 7C).

Fig. 7.

Podocyte-Specific Deletion of EGR1 Protects Against Podocyte and Glomerular Injury in PE Mice. A Schematic timeline illustrating the experimental procedure for generating podocyte-specific EGR1 knockout (EGR1^KO) mice. Pregnant mice were administered either PBS or L-NAME (100 mg/kg/day) intraperitoneally from E10.5 to E16.5. The podocyte-specific EGR1 knockout was achieved by crossing floxed EGR1 mice with Nphs2-Cre transgenic mice, resulting in the deletion of EGR1 specifically in podocytes. B Representative immunofluorescence images (left) and quantitative analysis (right) of glomerular EGR1 (red) and the podocyte marker synaptopodin (green) in kidney sections from control (Con) and L-NAME-treated mice. Nuclei are stained with DAPI (blue). n=6 per group; Scale bar, 50 µm. C Western blot and quantitative analysis showing the expression of EGR1 in primary podocytes from wild-type (WT) and EGR1^KO mice. D BCA protein quantification was used to assess urinary protein levels in WT and EGR1^KO mice treated with PBS or L-NAME. E Hematoxylin and eosin (HE) staining of kidney sections from WT and EGR1^KO mice with or without L-NAME treatment to visualize glomerular architecture such as mesangial expansion and capillary loop structure. F Immunofluorescence staining for podocyte-specific markers synaptopodin (green) and ZO-1 (red) in kidney sections from WT and EGR1^KO mice. G Western blot and quantitative analysis showing the expression of pro-inflammatory cytokines TNF-α and IL-1β in glomerulus from WT and EGR1^KO mice treated with L-NAME. H Immunofluorescence staining for F4/80 in kidney sections from WT and EGR1^KO mice. Data are presented as mean ± SD from 6 independent biological replicates; β-actin served as the loading control. Statistical significance was determined by Student’s t-test for panels B and C, and by two-way ANOVA for panels D, G, and H. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001

Compared with L-NAME-treated WT mice, L-NAME-treated EGR1^KO mice exhibited significantly reduced urinary protein excretion (Fig. 7D). Histological analysis, including HE staining (Fig. 7E), revealed better preserved glomerular architecture in L-NAME-treated EGR1^KO mice than in L-NAME-treated WT control mice. Specifically, L-NAME EGR1^KO mice demonstrated reduced mesangial expansion and maintained capillary loop structure as compared to L-NAME WT mice, indicating improved glomerular health and reduced glomerulosclerosis.

Immunofluorescence staining for the podocyte-specific structural proteins synaptopodin and ZO-1 (Fig. 7F) revealed increased expression in L-NAME EGR1^KO mice compared to L-NAME WT mice, further supporting the preservation of podocyte function and structural integrity. These results suggest that the deletion of EGR1 in podocytes helps maintain the critical components of the glomerular filtration barrier, which is disrupted in PE. Furthermore, western blot analysis (Fig. 7G) revealed markedly reduced levels of the proinflammatory cytokines IL-1β and TNF-α in the glomerulus of L-NAME EGR1^KO mice compared to L-NAME WT mice, suggesting that EGR1 deletion alleviates renal inflammation in the context of PE. This reduction in inflammatory cytokines underscores the pro-inflammatory role of EGR1 in podocytes and its contribution to renal protection in PE. Additionally, immunofluorescence staining for F4/80 (Fig. 7H) revealed a decrease in macrophage infiltration in L-NAME EGR1^KO mice compared to L-NAME WT mice, indicating reduced renal inflammation, which further reinforces the conclusion that EGR1-mediated inflammation plays a pivotal role in renal injury during PE.

Collectively, these findings confirm that EGR1 plays a pathogenic role in mediating podocyte injury in PE. Importantly, the targeted deletion of EGR1 in podocytes confers renoprotection, validating the functional relevance of the TMEM106A–EGR1 regulatory axis in the context of PE-associated glomerular injury. An integrative schematic model was constructed based on our findings (Fig. 8).

Fig. 8.

Schematic model depicting the proposed mechanism of EV-mediated podocyte injury in preeclampsia. TMEM106A levels were reduced in circulating plasma EVs from patients with PE. In cultured podocytes, exposure to PE-derived EVs downregulates TMEM106A expression, which subsequently upregulates EGR1, induces inflammatory responses and cytoskeletal remodeling, thereby leading to podocyte injury, slit diaphragm disruption, proteinuria, and renal damage of PE

Discussion

This study reveals a novel pathway through which reduced EV-derived TMEM106A induces podocyte injury in PE via activation of the transcription factor EGR1, thereby providing mechanistic insight into PE-associated renal impairment. Our findings demonstrate that TMEM106A, a transmembrane protein, plays a pivotal role in inflammation regulation and is significantly downregulated in plasma EVs from PE patients. This reduction in expression is related to the disorder of the cytoskeleton of podocytes, the elevation of inflammatory cytokines, and the increased urinary levels of nephrin and podocalyxin. In our study, podocytes exposed to plasma EVs from PE patients showed lower TMEM106A expression than those treated with control EVs, suggesting that the reduced level of TMEM106A carried by PE-derived EVs may be one mechanism contributing to TMEM106A downregulation in podocytes. Notably, we observed a negative correlation between urinary TMEM106A levels and podocyte injury markers in PE patients. Unlike systemic biomarkers such as soluble fms-like tyrosine kinase-1 (sFlt-1) [36], urinary TMEM106A reduction specifically mirrors podocyte injury in PE. Importantly, our study provides the first evidence supporting the diagnostic value of TMEM106A in PE.

Beyond its diagnostic value, our functional studies revealed that EV-derived TMEM106A deficiency directly promotes a proinflammatory state in podocytes. Accumulating evidence suggests that in chronic kidney disease, podocytes are induced by diverse pathological stimuli to secrete inflammatory factors, which not only aggravate podocyte injury through autocrine signaling but also promote disease progression by contributing to glomerulosclerosis and tubulointerstitial fibrosis [37, 38]. In line with these findings, our data reveal that EVs derived from PE plasma, characterized by low levels of TMEM106A, directly promote the release of elevated levels of inflammatory factors by podocytes. This suggests a vicious cycle in which PE-derived EVs not only damage podocytes but also trigger them to become a source of inflammation, potentially exacerbating renal pathology. However, important mechanistic questions remain. While our data indicate that the lack of TMEM106A in EVs is a key driver, we cannot exclude the possibility that the inflammatory cargo within these EVs directly stimulates podocyte cytokine production. Nevertheless, our podocyte-specific knockdown experiments demonstrated that reducing TMEM106A expression directly promotes inflammatory cytokine release and injury. These findings provide direct evidence supporting the functional role of TMEM106A deficiency in driving podocyte inflammation. However, given that altered sorting mechanisms in PE could enrich proinflammatory cargo in EVs [39, 40], the two mechanisms are not mutually exclusive and may operate synergistically. Therefore, future investigations should prioritize dissecting the specific cargo profile of TMEM106A-low EVs through proteomic and transcriptomic analyses.

The reduction in plasma EV-derived TMEM106A observed in PE may arise from interconnected disturbances at both the sorting and cellular expression levels. On the one hand, altered sorting and loading efficiency could be a direct cause. The biogenesis and cargo loading of EVs are highly regulated processes involving the endosomal sorting complex required for transport (ESCRT), lipid raft microdomains, and various RNA-binding proteins [17]. Placental hypoxia, a hallmark of PE, may disrupt the activity of these molecular machineries or alter membrane lipid composition, thereby specifically impairing the sorting of TMEM106A into EVs and leading to its relative deficiency within vesicles. On the other hand, PE placentas exhibit widespread epigenetic alterations, particularly dysregulated DNA methylation. Intriguingly, the expression of TMEM106A is regulated by the methylation status of its promoter region, with hypermethylation leading to its transcriptional silencing and reduced secretion [29, 41]. The hypermethylation landscape in PE placentas may thus directly suppress TMEM106A transcription in trophoblasts, reducing the protein pool available for packaging at the source. These pathways may synergistically contribute to the marked deficiency of TMEM106A within circulating EVs, a mechanism that requires further elucidation.

Importantly, transcriptomic sequencing of podocytes following TMEM106A knockdown revealed that EGR1 is a key downstream effector, and it is also a differentially expressed gene in the PAN-induced podocyte injury dataset (GSE124622). Mechanistically, TMEM106A deficiency results in the upregulation of EGR1 expression. EGR1 is a nuclear transcription factor that is rapidly and transiently induced by hypoxia, growth factors, and other cellular stimuli and regulates the expression of genes involved in cell growth, differentiation, proliferation, tumorigenesis, and inflammatory responses [42–48]. However, the role of EGR1 in podocyte injury in PE has not been previously investigated. Our data show that EGR1 overexpression aggravates podocyte cytoskeletal rearrangement and inflammatory responses. In murine models, both TMEM106A overexpression and podocyte-specific EGR1 deletion ameliorated renal and placental pathology in PE mice. Notably, TMEM106A restoration significantly increased fetal and placental weights. These findings are clinically relevant, as current standard therapies for PE, such as antihypertensive agents, effectively control maternal hypertension but show limited efficacy against fetal complications such as intrauterine growth restriction [49, 50]. Our results highlight the therapeutic potential of targeting the TMEM106A-EGR1 axis to address a critical unmet need in PE management. These results not only broaden the functional spectrum of TMEM106A beyond its known roles in tumor suppression and immune regulation but also position it as a critical mediator of PE-associated podocyte injury, whose modulation may offer dual benefits for both maternal renal function and fetal development. Although we have identified the central role of TMEM106A in PE-induced podocyte injury, the precise mechanism through which TMEM106A regulates EGR1 remains unclear. Whether TMEM106A directly suppresses EGR1 transcription or modulates its activity via other signaling pathways requires further investigation. In addition, future studies should include direct in vivo tracing of EV-derived TMEM106A delivery to podocytes, which would provide critical insights into its uptake, functional role, and regulatory impact on EGR1 signaling in the context of PE.

There are several limitations in our study. First, we did not directly isolate TMEM106A-positive and -negative EV subpopulations, as validated strategies for surface exposure and immunocapture of TMEM106A are currently unavailable. Consequently, we were unable to determine whether TMEM106A-negative EVs carry higher levels of pro-inflammatory factors, which could reveal whether TMEM106A-negative EVs represent a novel pro-inflammatory EV subpopulation. Second, our study focused specifically on the role of TMEM106A in podocyte injury under PE conditions and does not establish whether TMEM106A downregulation is a general indicator of podocyte injury across different disease models. Furthermore, we did not explore the broader relevance of TMEM106A regulation across various hypertensive settings. Future studies will be essential to address these questions and further clarify the role of TMEM106A in podocyte biology under diverse pathological conditions.

In conclusion, our work revealed that the reduction of EV-derived TMEM106A contributes to podocyte injury in PE by driving EGR1 upregulation. By elucidating this pathway, we provide mechanistic insights into the role of circulating EV-associated TMEM106A in PE-associated renal injury, offering a potential therapeutic avenue for mitigating renal dysfunction in PE.

Abbreviations

PE

Preeclampsia

TMEM106A

Transmembrane protein 106 A

EGR1

Early growth response 1

EV

Extracellular vesicle

PCX

Podocalyxin

eGFR

Estimated glomerular filtration rate

sFlt-1

Soluble Fms-like tyrosine kinase-1

DAMPs

Damage-associated molecular patterns

Authors’ Contributions

DZ, NS, XZ and MF conceived and designed the experiments. DZ, NS and XZ collected urine samples. ZZ and JS performed in vitro experiments. DZ, NS, XZ, SC, and XL performed in vivo experiments. SC and XZ conducted data analyses. MF provided suggestions and support. DZ, NS and MF drafted or revised the manuscript. All authors read and approved the final version.

Funding

None.

Data availability

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

Declarations

Ethics approval and consent to participate

The human study was approved by the Clinical Research Ethics Committee of The Second Hospital, Cheeloo College of Medicine, Shandong University (No. KYLL2024634), and the informed consent was signed by all participants in accordance with the Declaration of Helsinki. All animal experiments were approved by the Animal Care and Use Committee of The Second Hospital, Cheeloo College of Medicine, Shandong University (No. KYLL2024634).

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests