Early renal response to long term salt loading: Mitochondrial dysfunction, ER stress and uromodulin accumulation in the kidney medulla

Abstract

Kidneys play a critical role in maintaining water and electrolyte balance, but prolonged salt loading can disrupt renal function by inducing osmotic and oxidative stress. While high salt intake is well-known to contribute to hypertension and kidney damage, the early renal responses to mild, long-term salt intake, particularly in normotensive individuals, remain poorly understood. To help address this knowledge gap, we investigated the effects of exposing normotensive Wistar Kyoto (WKY) rats to 1% NaCl over a 3-month period, focusing on the medullary region and the adaptive cellular mechanisms in response to salt-induced stress. Additionally, we examined the acute effects of 4 hours of salt exposure on medullary tubules. The long-term salt intake did not significantly alter blood pressure or cause notable kidney damage, but did lead to differential expression of proteins associated with mitochondrial dysfunction and ER stress in the renal medulla. Acute 4-hour salt exposure triggered a rapid cellular response involving proteins linked to mitochondrial activity and oxidative stress responses. Both acute and chronic settings significantly reduced UMOD excretion with altered trafficking indicating intracellular accumulation within medullary cells. This provides evidence that chronic salt loading disrupts normal protein handling without immediate renal injury, shedding light on adaptive mechanisms in the kidney to mitigate osmotic stress. These early adaptations provide insight into the mechanisms underlying salt-related renal pathologies and may inform therapeutic strategies for individuals susceptible to the effects of dietary salt.

Introduction

Kidneys are among the most energy-consuming organs in the body due to their critical role in the reabsorption of water and electrolytes (1, 2). Long-term salt loading has significant detrimental effects on kidney function and health, primarily through mechanisms involving inflammation (3) oxidative stress (4) and altered metabolic processes (5). While the adverse effects of long-term salt loading on kidney health are well-documented, some studies suggest potential compensatory mechanisms, such as increased water intake and altered metabolic pathways, that may mitigate immediate damage but do not prevent long-term consequences. A growing body of evidence has linked high dietary salt intake to adverse renal outcomes, including mitochondrial dysfunction, oxidative stress, and impaired protein handling, which are implicated in the pathogenesis of hypertension and chronic kidney disease (CKD). However, the early renal responses to mild, long-term salt intake, particularly in the absence of cofounding factors such as hypertension, diabetes, or CKD, are less well understood.

In vivo studies into high salt diets often utilize concentrations such as 2%, 4% and 8% that exceed physiological levels or are conducted on salt-sensitive models (6–8). While such studies provide valuable insights, they may not fully represent the effects of typical dietary salt intake, raising questions about the generalizability of findings to human populations.

The renal medulla plays a crucial role in salt homeostasis, particularly in urine concentration and managing osmotic stress. Its naturally hyperosmotic environment makes it particularly susceptible to the challenges of increased salt intake. The renal medulla exhibits significant adaptive responses to chronic salt loading, primarily through osmotic regulation by altering gene expression of water regulatory molecules (9), homeostasis (10), and medullary adaptation mechanisms (11). Mechanisms underpinning this adaptation include changes in renal medullary blood flow and cellular plasticity. However, modest saline loading is not accompanied by renal medullary blood flow changes (11), indicating that other as yet unknown factors may be more critical in mediating the natriuretic response.

The thick ascending limb (TAL) of the loop of Henle controls the recycling of 30% of the urinary sodium, primarily via NKCC2. Uromodulin (UMOD), the most abundant protein in urine, has emerged as a key player in maintaining salt balance, particularly through its role in the TAL, where it modulates sodium reabsorption via NKCC2 transporter activity (12–15). We previously showed that salt-loading with 1% NaCl (hereafter 1% salt) for 3 weeks lowers urinary UMOD excretion in chronic hypertensive Stroke-Prone Spontaneously Hypertensive rats (SHRSP) (16).

This study aims to investigate the early renal responses to long-term salt loading in normotensive Wistar Kyoto (WKY) rats, focusing on the cellular mechanisms employed by medullary cells to adapt to the osmotic stress induced by prolonged sodium exposure. Specifically, we examine the impact of a 1% salt/NaCl as a saline drinking solution over a 3-month period on renal physiology, by exploring the proteomic changes in the renal medulla. The proteomics data highlighted mitochondrial dysfunction, oxidative stress, and ER stress, pathways essential for tubular health. UMOD, produced mainly by TAL and partly DCT cells, is the most abundant urinary protein and one of the most highly expressed kidney-specific transcripts. Its complex maturation requires extensive ER-dependent folding (24 disulfide bonds) and heavy glycosylation (17, 18). Therefore, salt-induced ER stress is predicted to disrupt UMOD processing and secretion, providing a functional link between the proteomic changes and tubular adaptation. To determine whether the effects of salt exposure on UMOD excretion are reversible, we also conducted an intermittent salt-loading study, where rats alternated between one week of 1% NaCl exposure and one week of water. Understanding these early adaptive mechanisms in a normotensive model could provide valuable insight into the progression of salt-related renal pathologies and inform therapeutic strategies for salt-sensitive individuals.

Methods

Animals and experimental design

All procedures were performed in accordance with Home Office regulation and the UK Animals Scientific Procedures Act 1986 (PPL No. PP0895181) and ARRIVE guidelines and were approved by the ethics review committee at University of Glasgow. We used the ARRIVE1 checklist when writing our report (19). Animals were housed under controlled environmental temperatures (21 ± 3°C) and lighting (12h light–dark cycles) and maintained on standard rat diet (rat no. 1 maintenance diet; Special Diet Services, Grangemouth, United Kingdom) and were provided tap water ad libitum. The chow contains a set amount of sodium and chloride, and each rat received a similar portion. Systolic blood pressure (SBP) was monitored weekly by tail-cuff plethysmography(20), in an operator-blinded fashion. We utilized 11 week-old Wistar-Kyoto (WKY) rats (referred to as ‘baseline’). Long-term salt study: At 12 weeks of age, littermates were randomized into groups of normal drinking water and salt-loaded (male n=6) over a period of 3 months, whereby the salt-loaded group received 1% NaCl/Salt/Saline solution (Sigma, Dorset, UK) in drinking water continuously during this period. Intermittent salt study: Another set of rats were assigned to group of normal drinking water (male, n=4) and 1% salt water (male, n=6) for 3 weeks, whereby the salt-loaded group would receive 1% salt intermittently in a 1 week on, 1 week off, and 1 week on pattern. Ex-vivo medullary tubule acute salt study: Separate two batches of WKY (n=3 per batch) without salt-loading were used to isolate medullary tubule. Power calculation: Sample size calculation was based on our previous study(16), which showed no significant blood pressure change or sex differences in WKY rats exposed to 1% NaCl. Using these data, we estimated a required sample size of 4; however, to enhance statistical strength, we used n=6 in the current study.

Cell line and culture conditions

Madin-Darby Canine Kidney (MDCK) cells stably expressing human uromodulin fused with green fluorescent protein (GFP) were established in LR’s laboratory (17). Cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) (ThermoFisher Scientific, Paisley, UK) supplemented with 10% fetal bovine serum (FBS, ThermoFisher Scientific, Paisley, UK) and 1% Penicillin/Streptomycin (ThermoFisher Scientific, Paisley, UK) at 37°C and 5% CO₂. The basal cell culture medium, DMEM, has a calculated osmolarity of 367 mOsm/L, with 220 mOsm/L contributed by 110 mM NaCl in the medium. After confluency, the cells were incubated in DMEM supplemented with 60mM salt or 120mM mannitol (as osmotic control) and incubated for 18h at 37°C, 5% CO₂. 60 mM NaCl or 120 mM mannitol contributes to approximately 120 mOsm/L, resulting in a final osmolarity of approximately 500 mOsm/L (170 mM NaCl). Media and cells were collected for uromodulin quantification.

Sample collection

Urine collection

Daily intake of water was recorded over 24 hours. Animals were individually housed in metabolic cages at baseline and once every experimental week for the intermittent salt study, or over 4 occasions (at baseline, 1-month, 2-month, and 3-month times points) for the long-term salt study, to estimate 24h urine output. Animals were acclimatized, 3 days before the first measurement. Twenty-four hour-urine samples were aliquoted and stored at −80°C until further use. Urine was not centrifuged prior to use, however debris contamination was not considered an issue as fecal matter and urine were separated via funnels in the metabolic cage. Aliquoted samples were not kept for longer than 3-6 months, in accordance with optimum UMOD storage conditions as set out by Youhanna et al. (21).

Blood collection

Heparinized blood was collected by tail vein puncture (baseline and each week for intermittent study or at 4 time points over 3 months for long-term salt study) under isoflurane anesthesia (2.5% isoflurane in 1.5L/min oxygen). At the end of both studies, heparinized and EDTA blood samples were collected by cardiac puncture and rats were killed by exsanguination under terminal general anesthesia. Biochemical plasma and urinary analyses for electrolyte, albumin and creatinine concentration were performed using Roche Cobas C311 Analyzer and commercially available rodent kits (Roche, Sussex, UK).

Tissue collection

Both kidneys were dissected from the rats and cut in half. One-half of one kidney was formalin-fixed and paraffin-embedded. The other half of the kidney snap-frozen in liquid nitrogen and stored at -80°C. The second kidney halves were individually dissected for cortex, medulla, and papilla regions and snap frozen in liquid nitrogen, later stored at -80°C until use.

Medullary tubule isolation and Ex-vivo salt incubation

Both kidneys from rats were used to isolate medullary tubules as described previously (22). Tubules were suspended into DMEM and incubated with either 300mOsm mannitol or NaCl (a final osmolarity of 667 mOsm/L (~33 mM NaCl), which falls within the physiological osmolarity range of the surrounding interstitial fluid in the renal medulla) for 4 hours at 37°C in a rotating incubator. After incubation, the tubule suspension was centrifuged at 1000 rpm for 10 mins, and the pellet was collected for protein extraction and labelling. This experiment was performed in two batches with n=3 rats per batch.

Protein extraction, quantification, and Western blot

Kidney tissues were lysed with Cell Extraction Buffer 5X PTR at 1X (Abcam, Cambridge, UK), whereas cells were lysed with RIPA lysis buffer for protein extraction. Both extraction buffers were supplemented with Octyl β-D-1-thioglucopyranoside, cOmplete™ Mini EDTA-free Protease Inhibitor Cocktail (Merck, Darmstadt, Germany), and PhosSTOP Inhibitor (Merck, Darmstadt, Germany). Membrane and cytosolic protein fractions from the medulla were extracted using Mem-PER™ plus membrane protein extraction kit (ThermoFisher Scientific, Paisley, UK) as per manufacturer’s instructions. Proteins were quantified via QuickStart Bradford assay (Bio-Rad, Laboratories Ltd, Hertfordshire, UK) according to manufacturer’s instructions. Proteins were separated on Bolt™ Bis-Tris plus gels (ThermoFisher Scientific, Paisley, UK). Semi-dry blotting (Power Blot, ThermoFisher Scientific, Paisley, UK) was performed on low fluorescence PVDF membrane (Merck, Darmstadt, Germany). Membranes were blocked with 5% BSA-PBST solution at RT for 1h, followed by overnight incubation with primary antibody at 4°C.

Primary antibody: anti-UMOD (1:1000, Abcam, Cambridge, UK), anti-vinculin (1:1000, Cell Signalling, Leiden, The Netherlands) anti-β-actin (1:1000, Sigma-Aldrich, UK), anti-BiP (1:1000, Cell Signalling, Leiden, The Netherlands), anti-PDI (1:1000, Cell Signalling, Leiden, The Netherlands), anti-flotillin 1 (1:5000, Abcam, Cambridge, UK), anti-flotillin 2 (1:2000, Abcam, Cambridge, UK), anti-calnexin (1:1000, Abcam, Cambridge, UK) and anti-epidermal growth factor (EGF) (1:1000, Abcam, Cambridge, UK).

Secondary antibody: anti-rabbit (1:10,000, Alexa Flour 800, ThermoFisher Scientific, UK), anti-mouse (1:10,000, Alexa fluor 680, ThermoFisher Scientific, Paisley, UK). In some cases, Recombinant Anti-UMOD antibody BSA and Azide free (1:800, Abcam, Cambridge, UK) was conjugated to 700 fluorophore using Alexa Fluor 700 Conjugation Kit Lightning Link (Abcam, Cambridge, UK) for imaging. Revert 700 total protein stain (LI-COR Biotechnology, Cambridge, UK) was used to analyse total protein for normalization for urine and cell culture media samples. Normalisation by sum of replicate was used when samples were in different blots, whereby all data points from a replicate are divided by the sum of all data points in that replicate (23).

Imaging was performed on Odyssey CLx and analysis on ImageStudio v5.0 (LI-COR Biotechnology, UK). In some cases, blots were stripped to remove prior antibodies with 0.2M NaOH for 10mins at RT, before blocking and re-incubation with another set of primary and secondary antibodies.

UMOD quantification

Urinary and serum UMOD concentration was quantified using rat ELISA kits (Abcam, Cambridge, UK) as per manufacturer’s guidelines for SimpleStep ELISA. The range of detection for UMOD standards was 625-40000 pg/mL for urine and 312.5-20000 pg/mL for plasma. Urine and plasma samples were diluted accordingly. All samples were quantified on the same day to minimize variation. Plates were read at 450 nm in a microplate reader (Victor Multilabel Plate Reader X3, Perkin Elmer Inc., Waltham, U.S.A).

Immunofluorescence

Paraffin-embedded kidneys were sectioned (5μm) for immunofluorescence. Unless indicated, all sections underwent a double immunohistochemistry protocol with a sequential antibody incubation. Antigen retrieval was performed on sections with hot citrate buffer (pH 6.0 at 95°C). The samples were then blocked in 1X Carbo-Free Blocking Solution (Vector Laboratories, Oxfordshire, UK) supplemented with 0.1% Tween-20 and incubated for 1h at RT. Samples were washed after blocking and in between antibody incubations with TBS containing 0.1% Tween-20. Sections were incubated with primary antibodies, followed by a wash, and then secondary antibody. All antibodies were diluted in 1X Carbo-Free Blocking Solution (Vector Laboratories, Oxfordshire, UK) supplemented with 0.1% Tween-20 and incubated for 1h at RT. Sections were mounted with Mounting Medium with DAPI – Aqueous, Fluoroshield (Abcam, Cambridge, UK).

For MDCK cells, cells were fixed in neutral buffered formalin (RT, 1 hour) after washes in PBS. Cells were then permeabilised with PBS containing 0.1% Triton X-100 for 10 minutes, before blocking in 1X Carbo-Free Blocking Solution (Vector Laboratories, Oxfordshire, UK) supplemented with 0.1% Tween-20 for 1h at RT. These were then directly mounted with Mounting Medium with DAPI – Aqueous, Fluoroshield (Abcam, Cambridge, UK).

The following primary antibodies goat anti-calnexin (1:100; Abcam, Cambridge, UK) and rabbit anti-UMOD (1:2000; Abcam, Cambridge, UK). Secondary antibodies used were goat anti-rabbit IgG (H+L) highly cross-adsorbed secondary antibody, Alexa Fluor 647 (1:500; ThermoFisher Scientific, Paisley, UK), donkey anti-goat IgG (H+L) cross adsorbed secondary antibody, and Alexa Fluor 546 (1:500; ThermoFisher Scientific, Paisley, UK).

Images were acquired with the Zeiss LSM 900 Axio Observer.Z1/7 confocal microscope equipped with a Plan-Apochromat 63x/1.40 Oil DIC M27 objective or Plan-Apochromat 40x/0.95 DIC M/N2. Operating with Zen Blue software version 3.3.89.0004 (Carl Zeiss, Oberkochen, Germany). Images were processed and analysed (including colocalization of UMOD and calnexin) using Zen Blue software. The analyst was blinded for imaging and analysis of sections.

Quantitative real-time PCR

Isolation of RNA from kidney and MDCK cells was performed using Qiazol and RNeasy mini spin kit (Qiagen, Manchester, UK) and the subsequent reverse transcription with High Capacity RNA to cDNA kit (Applied Biosystems, Leicestershire, UK). Gene expression assays were performed using TaqMan Fast Advanced Master Mix and species-specific TaqMan probes (ThermoFisher Scientific, Paisley, UK): Umod (Rn00567186), Havcr1 (Kim-1 - Rn00597703), Lcn2 (NGal - Rn00590612), Hprt (C102626258), and Ubc (Rn01789812). The expression levels were normalized to the housekeeping gene (Ubc for kidneys, Hprt for MDCK cells) and presented fold change to control conditions.

Endo H Deglycosylation

Endoglycosidase H (Endo H) assays were performed on medulla lysates dissected from kidneys of the animals, according to manufacturer’s instructions (Promega, Hampshire, UK). Briefly, samples were denatured at 95°C, followed by incubation with the Endo H at 37°C for 1 hour at RT. The resulting protein products were analysed by Western blotting. The Endo H enzyme cleaves the chitobiose core of high-mannose oligosaccharides and a limited number of hybrid oligosaccharides from asparagine-linked glycoproteins. It does not cleave complex glycans.

Proteomics and LC-MS/MS analysis

Kidney proteome of long-term salt study

Label free methodology was adapted for this study. Briefly, 100μg protein was digested after desalting with Amicon Ultra 3kDa MWCO (Millipore, USA) using 50 mM ammonium bicarbonate. Protein samples were subjected to heat denaturation (80°C), reduced using dithiothreitol for 30 min at 65°C and alkylated with iodoacetamide in the dark for 30 min at 25°C. Trypsin (Trypsin Gold, MS grade, Promega, USA) was added in 1:20 ratio trypsin:protein (w/w) and incubated overnight at 37°C. The reaction was stopped by the addition of 2 μL of formic acid.

Ex-vivo medullary tubule acute salt study

TMT-labelling methodology was used for this study as per manufactures guidelines (Tandem Mass Tag™ 6-plex kit, ThermoFisher, UK). Briefly, 50μg protein was digested as mentioned above. The digested peptides were labelled using the TMT label reagent in ratios recommended by the manufacturer. The labelled samples were fractionated using Pierce™ High pH Reversed-Phase Peptide Fractionation Kit (ThermoFisher, UK).

Nano LC MS/MS analysis and quantification

LC-MS/MS analysis was performed using an UltiMate 3000 nano-flow system (Dionex/LC Packings, USA) connected to a LTQ Orbitrap Velos FTMS hybrid mass spectrometer (Thermo Fisher Scientific, Germany) with a nano-electrospray ion source. A 20 μL sample was loaded onto a Dionex 0.1 × 20 mm, 5 μm C18 nano trap column at a flow rate of 5 μL/min, using a mobile phase of 98% water, 0.1% formic acid, and 2% acetonitrile. The sample was then eluted onto an Acclaim PepMap C18 nano column (75 μm × 50 cm, 2 μm, 100 Å) at a flow rate of 0.3 μL/min. Both the trap and nano columns were maintained at 35°C. Elution was achieved using a linear gradient between solvent A (0.1% formic acid in 98:2 acetonitrile/water) and solvent B (0.1% formic acid in 20:80 acetonitrile/water). The sample was ionized in positive ion mode using a Proxeon nano-spray ESI source (Thermo Fisher Hemel, UK) and analyzed by the Orbitrap Velos FTMS (Thermo Finnigan, Bremen, Germany). The column was washed and re-equilibrated between injections. Ionization voltage was set at 2.8 kV, and the capillary temperature was maintained at 250°C. The mass spectrometer operated in data-dependent MS/MS mode, scanning from 380 to 1600 amu. The top 20 multiply charged ions from each full scan were selected for MS/MS fragmentation using HCD at 40% collision energy. The resolution was set to 60,000 for MS1 and 7,500 for HCD MS2 analysis. MS raw data files were searched against the Uniprot rat database using MaxQuant v1.6.17.0 (24). The statistical analysis was performed using Perseus v 1.6.15.0 (25). Gene ontology was performed using ShinyGO V0.76 (26). Pathway analysis and networks were generated in Qiagen Ingenuity Pathway Analysis (27)

Statistics

Statistical analysis was performed on GraphPad Prism version 9 (GraphPad Software, California, USA). Student’s t-test, Welch’s t-test or Mixed-effects analysis was performed as appropriate. The effect of two factors was tested by two-way ANOVA (or mixed model). All statistical tests were two-tailed and P-value <0.05 was considered significant. For ELISA analysis, four-parameter curve fit (4PL) without constraints was used to determine the curve fit for standard values in GraphPad Prism. For PCR, the Ct values obtained from QuantStudio 12K Flex software version 1.3 (ThermoFisher Scientific, Paisley, UK) were used for calculation of ΔCt and fold change (FC) manually in Excel, and the graphs were created in GraphPad.

Results

Long term salt loading does not cause major fluctuations in blood pressure and kidney damage

To explore the effects of long-term salt-loading on renal excretion and physiology, we characterised parameters relevant to Na⁺ and water homeostasis in the WKY rats. The salt-loaded rats were provided with 1% salt which equivalates to 1.22g of NaCl per day or 0.48g of sodium per day (considering average fluid intake was 122mL/day). The rats receiving long-term salt loading secreted 0.463g of NaCl per day in their urine. We did not measure accumulation in the skin and other organs (28) or secretion via faeces (~10% of sodium is excreted via the faeces (29)) as we focused on urinary parameters. To translate this to humans, several blood pressure guidelines recommend low sodium intake (<2.3g of sodium intake, or 5.8g/day of NaCl) (30). Long-term salt-loading did not significant change SBP or body weight in these animals (Figure 1A and Supplemental Figure S1A). However, as expected, salt-loading did lead to increased urinary output, fluid intake, and urinary sodium excretion (Figure 1B-D), but plasma sodium levels were not altered significantly (Supplemental Table S1). We explored the expression of injury markers kidney injury marker-1 (KIM-1; proximal tubule) and neutrophil gelatinase-associated lipocalin (NGAL; TAL and distal tubule) in long-term salt-loaded WKY rats. The absence of any significant change in levels (Figure 1E and F) provides evidence for no major kidney damage, also supported by the lack of change in total kidney weight (Supplemental Figure S1B).

Figure 1. Long-term salt-loading increases urinary output and sodium excretion in WKY.

Systolic Blood Pressure (SBP) was assessed by tail -cuff measurements. (A) shows the difference in SBP in WKY from baseline measurements in control (n=6) and 3-month (n=6) salt-loaded groups. (B) Urinary output and (C) fluid intake of WKY rats were significantly increased during 3 months of salt-loading. Data represent mean ± S.E.M. **P<0.01, ***P<0.001, ****P<0.0001 (Mixed-effects analysis with Šídák's multiple comparisons test). (D) Urinary sodium excretion (u-sodium) was significantly increased in the salt-loaded group, represented as the delta change from baseline values for control and 3-month salt-loaded groups (n=6) (± S.E.M). ***P<0.001 (Unpaired t-test). Gene expression represented as fold change to control ± S.E.M of (E) kidney injury marker-1 (KIM-1) and (F) neutrophil gelatinase-associated lipocalin (NGAL), n=6 per group. mRNA levels were normalised to Ubiquitin C (Ubc) expression.

Kidney proteome changes with 1% salt incubation over long-term

To create insight into any molecular changes chronic salt loading causes in the kidney, we undertook a proteomic analysis of the kidney (Figure 2A, B). We identified 107 differentially expressed proteins (DEPs) between control and long-term salt-loaded WKY (Figure 2C, Supplemental Table S4). This was a 9% protein change observed in the identified proteome (total proteins identified after filtering 1187). IPA canonical pathway analysis identified these DEPs to be enriched to pathways such as mitochondrial dysfunction, glutathione-mediated detoxification, EIF2 signalling pathway, neutrophil extracellular trap pathway and NRF2-mediated oxidative stress (Figure 2D, Supplemental Figure S2). The DEPs overlapping in these canonical pathways were CYCS, FIS1, GSTA4, GSTK1, GSTT1, MT-ND5, NDUFV2, TOMM70, GSTA2, EIF3E, MAP2K1, RPL13A, RPL15, RPL17, RPL23, RPL3, RPS11, RPS2, DNAJA2, DNAJC3, ERP29 (Supplemental Figure S2). The majority of DEPs belonged to mitochondria and gene ontology enrichment provided further evidence of potential mitochondrial dysfunction (Supplemental Table S2). These changes in protein expression indicate a complex adaptive response to salt loading in the kidney in the absence of overt kidney injury.

Figure 2. Long-term salt-loading changes total kidney proteome.

(A) Schematic diagram showing the study. (B) Principal component analysis (PCA) displays distribution of the biological replicates from control (blue) and salt (pink) kidney proteome. (C) Volcano plot showing up- (red) and down- regulated (blue) proteins in long-term salt-loaded WKY. Top 20 differentially expressed proteins are labelled. (D) IPA-canonical pathway analysis of differentially expressed proteins shows mitochondrial dysfunction, glutathione-mediated detoxification, EIF2 signalling pathway and NRF2-mediated oxidative stress pathways.

Acute changes in kidney medulla with high salt

The medulla contains the loop of Henle and the collecting ducts, which are critical for urine concentration. The countercurrent multiplication system in the loop of Henle creates a gradient of increasing osmolarity, enabling the medulla to reabsorb water and concentrate urine effectively, especially under conditions of osmotic stress. To shed light on acute medullary molecular responses to salt, we performed a medullary tubule extraction from WKY rats and subjected them to 300 mOsm NaCl (Salt) ex-vivo for 4 hours (Figure 3A). As a control, we incubated the tubules with 300mOsm of Mannitol. Proteome analysis of these ex-vivo tubules detected 2044 high confidence proteins between mannitol tubule and salt tubule groups (Supplemental Table S4). Among this, 137 proteins were significantly differentially expressed with 45 upregulated and 92 downregulated in salt-tubule compared to mannitol-tubule group (Figure 3B). Gene ontology analysis of DEPS in salt-loaded medullary tubules after 4 hours of exposure shows a rapid cellular response to osmotic stress. Key changes include enhanced protein synthesis, gene expression regulation, intracellular reorganization, and adjustments in cellular metabolism and energy production (Figure 3C). Other major changes included the upregulation of ATP synthase subunits and oxidative phosphorylation components (Figure 3D), and the various ion channels and transporters involved in ion homeostasis (Figure 3E). Mitochondrial dysfunction, oxidative stress and ER stress were commonly enriched pathways in the kidney proteome under long-term salt loading and the kidney medulla proteome under acute salt loading (Supplemental Figure S2). Although the proteins identified in these pathways are not common.

Figure 3. High salt-loading causes acute change in medulla–

(A) Medullary tubules were extracted from WKY and ex-vivo incubated with 300mOsm NaCl and mannitol for 4 hours. The panel depicts the methodology for TMT-labelled proteomics. (B) Volcano plot showing distribution of significantly up-(red) and downregulated (blue) proteins in salt group. Top 20 differentially expressed proteins are labelled. (C) Gene ontology analysis showed key changes in medullary tubules after 4 hours of salt incubation. (D) Violin plot showing the differentially expressed proteins belonging to mitochondrial dysfunction. Data represent mean ± S.E.M. **P<0.01, ***P<0.001, ****P<0.0001. (E) Heatmap of various channels and transporters involved in ion homeostasis.

Long-term salt-loading induces ER stress in medulla

High salt intake can simultaneously induce ER stress and mitochondrial dysfunction via osmotic and oxidative stress (31, 32). To evaluate the impact of long-term salt-loading on the ER environment in the total kidneys (cortex and medulla) of the long-term salt-loaded rats, we examined the levels of ER stress markers Binding Immunoglobulin Protein (BiP) and Protein Disulphide Isomerase (PDI), which revealed no statistically significant difference (Figure 4A-B). However, BiP and PDI showed an increase in protein expression in the medulla of 3 months salt-loaded group (Figure 4C and D) supporting ER stress in the medulla. To determine if the increase in ER stress marker levels is due to prolonged salt exposure, we compared BiP levels between the long-term salt study and the 3-week acute salt study (rats from our previous study (16)). This revealed BiP was only upregulated in long-term salt exposure (Supplemental Figure S3), supporting activation of ER stress in the medulla but only occurs after longer exposure to salt.

Figure 4. The medullary lysate was extracted from total kidneys of rats from the long-term salt study.

UMOD protein levels and representative Western blots for Binding Immunoglobulin Protein (BiP) and Protein Disulphide Isomerase (PDI) for total kidney lysates (A, B), and BiP and PDI for the medulla (C, D). Error bars represent mean ± S.E.M., *P<0.05 (Welch’s t test). n=6 kidney (per group) was evaluated.

Given the evidence of ER stress induced by prolonged salt exposure, we next aimed to investigate the impact on renal function, specifically medullary function, through the assessment of uromodulin levels.

Salt decreases UMOD excretion rate

UMOD regulates electrolyte balance, manages osmotic stress, and responds to kidney stressors in the renal medulla, making it a key marker for assessing medullary function. We previously showed that in WKY animals on 1% salt, UMOD excretion rate decreased within 3 days (16). However, the temporal relationship between UMOD levels and salt loading remains poorly defied. Thus, we tested whether increasing the duration of salt-loading affects urinary UMOD excretion rate and expression. This revealed that long-term salt exposure in WKY rats significantly attenuated urinary UMOD excretion rate as levels remained lower than baseline at 1-month, 2-month, and 3-month timepoints (Figure 5A). After 3 months of salt-loading, there were no significant fluctuations in serum UMOD levels (Figure 5B). Thus UMOD levels respond quickly to salt exposure without any overt compensation when rats are exposed to long-term salt.

Figure 5. Long-term salt-loading decreases urinary UMOD excretion in WKY.

Urine samples were collected over 24 hours over 4-time points in the study using metabolic cages, whereas blood was collected at the end of the study by cardiac puncture. Urinary and serum UMOD concentrations were quantified by ELISA. Urinary UMOD was normalised to 24-hour output for urine and represented as excretion rate (mg/h). (A) represents urinary UMOD excretion rate for the long-term salt study (n=6 control, n=6 salt). Data is shown as mean ± S.E.M. **P<0.01, ****P<0.0001 (Mixed-effects analysis). (B) shows plasma UMOD levels after 3 months in animals of the long-term salt study are unchanged (n=6 control, n=6 salt), suggesting salt is directly affecting urinary UMOD excretion. Data is shown as mean ± S.E.M. (C) represents urinary UMOD excretion rate for the intermittent salt study (n=4 control, n=6 salt). Data is shown as mean ± S.E.M. **P<0.01, ****P<0.0001 (Mixed-effects analysis).

To test whether the effect of salt on UMOD excretion was reversible, we next conducted an intermittent salt-loading study (SBP, weight, urine output and other characteristics and biochemistry detailed in Supplemental Figure S4 and Table S3). Analysis of the UMOD urinary excretion rate showed lower levels of UMOD with salt-loading at all time points when compared to controls (Figure 5C). However, the UMOD excretion rate did increase in the salt-off periods relative to the salt-on periods, although it did not reach control levels. This supports that the effect of salt exposure on UMOD excretion is reversible.

To confirm whether this effect of salt is specific to UMOD, we analysed the levels of another medullary glycosylated protein, Epidermal Growth Factor (EGF), which is excreted by the TAL. EGF excretion into the urine (Supplemental Figure S5A) and protein levels in the kidney (Supplemental Figure S5B) did not alter significantly, providing evidence that salt is influencing UMOD directly and acutely at the excretion level.

Prolonged salt-loading increases ER accumulation of UMOD

Given the consistent reduction in UMOD excretion during long-term salt-loading (Figure 5A-B), we next investigated its impact on UMOD trafficking. We first determined UMOD mRNA and protein levels in total kidney extracts of 3-month salt-loaded normotensive WKY rats to assess any changes in transcriptional and post-transcriptional regulation of UMOD. The combination of unaltered mRNA level with elevated total kidney UMOD protein levels were in salt-loaded groups support that chronic salt loading acts post-transcriptionally (Figure 6A and B). In comparison, intermittent salt-loading did not affect UMOD at either level (Figure 6C and D). These results support that the slower UMOD excretion rate in long-term salt-loaded animals likely reflects intracellular retention of UMOD.

Figure 6. Comparison of UMOD in long-term salt loading versus Intermittent salt loading.

(A) Total kidney (cortex and medulla) UMOD mRNA expression of long-term salt-loaded group n=6 control, n=6 salt. mRNA levels were normalised to Ubiquitin C (Ubc) expression. (B) Total kidney UMOD protein levels and representative Western blot of each group in long-term study (n=5 control, n=5 salt), showing increased UMOD and intracellular accumulation. Data represented as mean ± S.E.M. *P<0.05 (Welch’s t-test). (C) Total kidney UMOD mRNA expression of intermittent salt-loaded group n=4 control, n=6 salt. mRNA levels were normalised to Ubiquitin C (Ubc) expression. (D) Total kidney UMOD protein levels and representative Western blot of each group of the intermittent study (n=4 control, n=6 salt). (E) UMOD protein levels and representative Western blots from control (n=5) and salt-loaded groups (n=5) of the long-term study in the medullary fraction of their kidneys. (F) Total UMOD protein levels and representative Western blots from control (n=4) and salt-loaded groups (n=6) from the intermittent study in the medullary fraction of their kidneys. Data represented as mean ± S.E.M. The second lower molecular weight band in the Western blots for UMOD is non-specific.

To interrogate any effects of long-term salt-loading on UMOD trafficking we performed subcellular fractionation analyses of isolated medullary regions. The medulla showed upregulated total medullary UMOD protein levels, unlike the intermittent salt study (Figure 6E and F). We found that 3 months of salt-loading revealed significantly increased UMOD protein levels in the cytoplasmic fraction, unlike the membrane fraction which remained unchanged (Figure 7A and B). Since UMOD is being retained intracellularly, and the rate-limiting step in UMOD secretion is its release from the ER, we looked at UMOD levels in the ER. We performed co-localization immunofluorescence analysis using calnexin as an ER marker to understand the distribution of UMOD in TAL tubules of the kidney. This showed increased ER accumulation of UMOD in the salt-loaded WKY (Figure 7C). This was further supported by the fact that the ER (immature) form of UMOD, consisting of simple immature N-glycans, was significantly increased in the salt-loaded group compared to the membrane (mature) form (Endo H-resistant) which contains complex N-glycans, as demonstrated by the Endo H assay (Figure 7D). The significant UMOD accumulation in the ER of the long-term salt-loaded animals, implicates that salt regulates the maturation of UMOD in TAL cells and leads to its retention if continued for longer periods of time. In contrast, intermittent salt-loading had no significant effect on the subcellular localisation of UMOD (Figure 7E and F) in the kidney since they were collected in the salt-off phase of the study.

Figure 7. Long-term salt-loading induces accumulation of immature UMOD in the ER of medullary TAL in normotensive WKY rats.

Kidneys of control and salt-loaded animals (n=6, per group) from the long-term study were isolated. (A) and cytosolic fractions (B) isolated from the medullary region of the control (n=6) and salt-loaded groups (n=6). Data represented as mean ± S.E.M. The plasma membrane (referred to as membrane fraction) was validated by the presence of membrane marker Flotillin-1. The cytoplasmic fraction contains the contents of all intracellular organelles (e.g. ER, Golgi apparatus, etc.) and the cytoplasm as confirmed by presence of ER chaperone calnexin, and Flotillin-2 as a marker of the cytosol. Vinculin was used as a loading control. *P<0.05, and **P<0.01 (Welch’s t test). (C) Representative immunofluorescence analysis showing UMOD (green) and ER marker calnexin (red) in from medulla of rat kidney sections. Co-localization of UMOD and calnexin is represented by yellow. Nuclei are stained in blue with DAPI. Scale bar represents 20 μm. The scatter plot on the right represents Pearson’s correlation coefficients which were calculated for four representative tubules (ROI, regions of interest) from 5 samples per group. Error bars represent mean ± S.E.M. **P<0.01 (Welch’s t test). Images were evaluated in a blinded fashion. (D) Endo H assay of the medulla samples from the control (n=6) and long-term salt-loaded (n=6) animals. Western blot shows representative bands of the endo H-resistant UMOD with complex glycans (mature, top band represented by “*”) and the ER form of UMOD (immature, bottom band). Error bars represent mean ± S.E.M. *P<0.05. (Welch’s t test). Western blots for UMOD from subcellular fractions of the medulla isolated from the intermittent salt-loading study, showing (E) membrane fraction and (F) the cytosolic fraction. Control n=4 and salt n=6. Data represented as mean ± S.E.M.

Salt lowers UMOD excretion in UMOD-expressing MDCK cells

To confirm the direct effects of salt on UMOD trafficking and extend the in vivo data, we utilized an in vitro model constitutively expressing human UMOD-GFP in MDCK cells. Cells were incubated with 60mM salt (in addition to existing 110mM in basal media) for 18 hours to model short-term exposure unlike the long-term salt-loading in rats. Cells exhibited a decrease in UMOD in the media and no change in cell lysate, as well as mRNA levels, directly indicating a reduction in UMOD secretion (Figure 8A-C). This was further corroborated by immunofluorescence analysis of UMOD, which showed significantly lower levels of UMOD on the surface of the cells, at the membrane in the salt-incubated cells compared to mannitol controls (Figure 8D, E, and F). Thus, these data establish that chronic salt loading lowers UMOD secretion via causing its ER retention and ER stress.

Figure 8. The proportion of UMOD at the membrane is decreased after salt incubation in MDCK cells.

Human UMOD-GFP expressing MDCK cells were incubated with DMEM media with additional 60mM salt or 120mM mannitol (osmotic control) for 18 hours. (A) Secreted UMOD protein levels in the media and representative Western blots. Media samples were normalised to total protein stain revert 700 (B) UMOD protein levels in the cell lysate and representative Western blots. UMOD was normalised to Vinculin. (C) mRNA expression shown as fold change relative to mean of control ± S.E.M and normalised to HPRT expression. (n=5 mannitol, n=5 salt). Data is shown as mean ± S.E.M. (D) Immunofluorescence analysis of UMOD in UMOD-GFP expressing MDCK cells after incubation with 60mM salt or 120mM mannitol as an osmotic control for 18 hours at 40X magnification. Representative images of non-permeabilized and permeabilized cells with UMOD (green) are shown and associated relative fluorescence units (RFU) (non-permeabilized (E) and permeabilized (F) calculated from the images displayed as scatter plots. White dashed boxes and arrows represent zoomed in regions of interest. Data is represented as mean ± S.E.M (n=6 images per group) **P<0.01(Welch’s t-test). Scale bar represents 20μm. DAPI used as a nuclear stain (blue). Images were evaluated in a blinded fashion

Discussion

Using Wistar Kyoto rats, we aimed to examine the effects of long-term salt-loading on renal physiology and excretory function and the underlying molecular mechanisms employed by medullary cells to adapt to the osmotic stress induced by prolonged salt intake. Our findings indicate that a 1% salt intake over a 3-month period in WKY rats did not cause a significant alteration in systolic blood pressure or notable kidney damage. However, it did activate a molecular response linked to mitochondrial dysfunction and oxidative stress. In addition, ER stress was observed in the renal medulla with altered UMOD trafficking leading to its accumulation in the ER of epithelial cells in the TAL. This results in reduced excretion rate of UMOD. Even an acute short term exposure to salt elicits a profound molecular response in medullary tubules encompassing differential regulation of proteins involved in mitochondrial function and oxidative stress. Additionally, the rate of UMOD excretion in response to acute and intermittent salt-loading remains reduced, mirroring the effects observed during long-term salt exposure, and this reduced excretion is driven by sodium levels rather than osmotic pressure. These findings support that long-term salt-loading induces osmotic and oxidative stress by disrupting ER and mitochondrial function, with a reduction in UMOD excretion rate emerging as one of the primary and early renal responses to salt-induced stress.

The absence of significant blood pressure changes and kidney damage indicates that WKY rats can tolerate long-term salt intake without immediate overt adverse effects. This finding is crucial as it suggests that the kidneys effectively manage excess sodium, maintaining overall homeostasis. However, the increased urinary output and sodium excretion reflects the kidneys' adaptive response to high salt intake, emphasizing the importance of efficient renal function in mitigating potential hypertensive effects. Proteomic changes highlight the kidneys’ adaptive mechanisms to prolonged salt exposure, reflecting a complex response to increased osmotic and oxidative stress. The upregulation of proteins involved in mitochondrial function and oxidative phosphorylation indicates heightened metabolic demands and a need for efficient energy production. Concurrently, the overexpression of glutathione S-transferase isoforms (A6, alpha-4, kappa 1, theta-1) suggests enhanced detoxification and protection against oxidative damage. In contrast, the downregulation of proteins related to energy production, lipid metabolism, and oxidative stress suggests a shift towards protective strategies aimed at minimizing cellular damage. Overall, these changes underscore the kidney's effort to preserve cellular homeostasis and adapt to the increased osmotic and metabolic challenges.

The induction of ER stress in the medulla after long-term salt-loading is significant as it points to localized stress responses. The medulla's critical role in urine concentration and handling osmotic stress makes it particularly vulnerable to prolonged high salt intake. Differentially expressed proteins in salt-loaded medullary tubules after 4 hours of exposure revealed a rapid and coordinated cellular response to osmotic stress. Enhanced protein synthesis and gene expression regulation were indicated by significant enrichment in cytoplasmic translation and RNA binding processes. Additionally, substantial intracellular reorganization and reinforcement of cellular integrity were reflected in changes in the cytoplasm, ribosomal subunits, mitochondria, and cell junctions. Metabolic and energy production adjustments were highlighted by increased activity in cellular metabolic processes and mitochondrial components. The upregulation of various ATP synthase subunits (Atp5 series) and electron transport chain components (Cox7a2, Mtco1, Mtco2, Mtnd1, Mtnd2, Mtnd4, Mtnd5, Ndufs4) suggests that high salt conditions increase the demand for ATP and enhance mitochondrial oxidative phosphorylation. Cells also activated oxidative stress management pathways and regulated ion homeostasis, as evidenced by enriched processes related to oxidative stress response and potassium and sodium ion homeostasis. Downregulation of Gsta1 and Sod1 reflect changes in the cell’s detoxification pathways and oxidative stress responses, possibly indicating a shift in how cells manage reactive oxygen species and other toxic by-products under high salt conditions. Upregulation of the ATPase subunits (Atp1a1, Atp1b1) and Fxyd2 indicates an increased activity in ion transport mechanisms, crucial for maintaining cellular ion homeostasis under high salt conditions. While, Slc9a3r1 is downregulated, potentially reflecting a shift in sodium reabsorption mechanisms. Most other transporters and channels show no change in expression, indicating stable handling of various ions and nutrients. It is important to note that these results represent changes in protein expression and do not necessarily reflect the functional activity of these channels and transporters.

The observed reduction in UMOD excretion and its intracellular retention suggest that long-term salt-loading disrupts protein trafficking, potentially impairing renal function over time. Higher salt intake leads to increased reabsorption of salt in TAL via NKCC2 activity, and UMOD regulates this activity (15). Most of the preclinical studies were performed with UMOD deficient or overexpressing mice, which therefore do not necessarily recapitulate the human disease. Here we demonstrate in normotensive WKY that UMOD excretion is directly and acutely lowered after 1 week of salt-loading and remains low after 3 months of continuous salt-loading. In a separate study with hepsin-deficient mice, high-salt exposure (2% salt) over 2 months increased urinary UMOD excretion in wildtype C57/BL6 mice, and due to defective UMOD processing in hepsin-deficient mice, leads to increased intracellular accumulation of UMOD, ER stress and tubular injury (33). However, in our previous study, we observed a significant reduction in UMOD excretion after 3 weeks of salt-loading in SHRSP rats (16). Interestingly, UMOD excretion rate in the salt group of the intermittent salt study as well as long-term salt study showed a 66.6% reduction in the UMOD excretion rate. It is reasonable to postulate that a daily 66.6% reduction of the most abundant protein in urine will significantly affect the normal physiology of kidney. Alternatively, it is reasonable to assume that the reduction of urinary UMOD may also be part of a negative feedback loop, to reduce hyperosmolality in the urine as a result of salt-loading and high concentrations of NaCl in the urine.

The intermittent salt study proves that the first point of disturbance of salt in UMOD trafficking is during its excretion, whereby UMOD excretion levels were lowered after 1 week of salt-loading. This is a specific effect on UMOD itself in the TAL, and not on other proteins, as the secretion of EGF (a protein also secreted by the TAL (21)) was unaffected. Also, given that UMOD secretion is influenced by salt in our MDCK cell line, and this cell line does not express NKCC2, the involvement of NKCC2 in increasing the retention of UMOD in the ER is unlikely. It is most likely a reversible homeostatic effect, given the fact that UMOD excretion increased after removal of salt-loading. It is unlikely that osmolarity is affecting the UMOD secretion in these cells, given that osmolarity was controlled for and kept the same between experiments using mannitol. The quick action of salt on UMOD secretion in these cells, with a difference found after 18 hours, could reflect the importance of sodium handling by cells and the role UMOD plays in this. Nonetheless, it is logical to conclude that the reduction in UMOD excretion seen here in a normotensive setting is due to influence of salt on the cells of the TAL and occurs irrespective of a blood pressure or kidney injury factors, bolstering our previous findings (16). It is interesting that salt has a unilateral effect on UMOD trafficking (i.e. an apical trafficking rather than a basolateral), given that serum UMOD levels were unaffected. It may also suggest a different trafficking pathway for basolateral secretion of circulating UMOD. Circulating UMOD has a consistent inverse relationship with hypertension (34), it may be hypothesised that circulating UMOD excretion will only be affected in a pathophysiological circumstance (e.g. with kidney injury) as a protective measure, which is not the case here.

Previous studies reported that a high-salt diet induces mitochondrial dysfunction in Dahl salt-sensitive rats and causes kidney damage (35). Moreover, high dietary salt intake increases vascular oxidative stress and contributes to renal damage via increased inflammation (36–38). A study with C57BL/6J mice indicates that high salt loading associated with oxidative stress by downregulating Nrf2 expression in collecting duct of kidney (4). Since sodium loading influences the pH and redox potential, and these are critical in the correct folding of proteins in the ER (39), the long-term salt-loading is likely contributing to the misfolding of UMOD and causes ER stress which is a key aspect of renal damage in hypertension (40). Thus, an interesting hypothesis is that these stress signals may transfer to mitochondria through a contact point between ER and mitochondria known as mitochondria-associated membranes (MAMs), via UPR activation. (41, 42). Some studies indicate the metabolic events between ER and mitochondria, and their occurrence increase under stress conditions, specifically during the early stage of ER stress (43–45).

This study has a few limitations. Firstly, the molecular signalling pathways influenced by salt that regulate UMOD trafficking are still unclear. Clearly, more mechanistic insight into the interactions between salt and UMOD trafficking are required. Moreover, we have not proven that the long-term effects of salt-loading are reversible and whether the ER accumulated UMOD can be rescued. This is certainly possible with short-term exposure. However, while we found salt-induced functional changes in the form of attenuated UMOD excretion, there are no obvious renal structural changes, highlighting the dynamic regulation of UMOD in response to salt. In this study, we provided 1% NaCl as a saline drinking solution rather than modifying dietary salt intake. While this approach allows for precise control of sodium exposure and aligns with previous studies on renal sodium handling, it does not fully replicate typical human salt consumption, which primarily comes from dietary sources. Notably, the 'only saline' drinking paradigm has been used in the literature to induce osmotic stress and dehydration responses. However, in our study, the small but significant increase in plasma sodium and chloride without changes in hematocrit suggests that any effects on fluid and volume homeostasis were mild. These factors should be considered when interpreting the translational relevance of our findings. We acknowledge as a limitation that direct measurements of plasma and urine osmolality in the rat studies were not performed. However, osmolality was carefully accounted for in our in vitro and ex vivo experiments through the use of mannitol controls, allowing us to isolate the specific effects of sodium on UMOD accumulation and ER stress.

It should be stated that this study does not serve to analyse which model is most suitable for studying salt and UMOD. Moreover, the focus is not on the disease (such as hypertension, CKD) itself, but rather the mechanisms that lead to it under sustained over salt consumption. Key to the present study is the use of 24-hour urine as a golden standard for analysing UMOD excretion rates. This removes the need for normalisation to creatinine, as is required for spot urine samples, as was used in other studies of UMOD(46, 47). We exclusively use male rats here, as our previous study showed no discernible sex differences in UMOD excretion rate in response to salt-loading (16).

This study highlights the progressive effect of salt on UMOD and renal physiology in a normotensive background. Only after prolonged salt exposure, and continued ER accumulation, is an ER stress response triggered. This is one of many possible mechanisms of effects/damages caused by long exposure of salt even at a physiological level. Understanding the pathways involved in interactions of salt and UMOD would provide novel therapeutic interventions for hypertensive patients with UMOD risk variants, as they may be more susceptible to environmental stressors such as salt.

News and Noteworthy.

This study reveals that even in normotensive Wistar Kyoto rats, mild long-term salt loading induces early renal stress without overt kidney damage or hypertension. Novel findings include reduced uromodulin (UMOD) excretion and altered intracellular trafficking in the renal medulla, alongside mitochondrial dysfunction and endoplasmic reticulum stress. These data highlights UMOD as a sensitive marker of salt-induced renal adaptation and provide insight into early cellular responses to salt before clinical disease onset.

Data Availability

All supporting data are included within the main article and Supplemental data. Further enquiries can be directed to the corresponding author.