Role of paraoxonase 3 in regulating ENaC-mediated Na⁺ transport in the distal nephron

Abstract

Paraoxonase 3 (PON3) is expressed in the aldosterone-sensitive distal nephron (ASDN), where filtered Na⁺ is reabsorbed mainly via the epithelial Na⁺ channel (ENaC) and Na⁺-coupled co-transporters. We previously showed that PON3 negatively regulates ENaC through a chaperone mechanism. The present study aimed to determine the physiological role of PON3 in renal Na⁺ and K⁺ homeostasis. Pon3 knockout (KO) mice had higher amiloride-induced natriuresis and lower plasma [K⁺] at baseline. Single channel recordings in split-open tubules showed that the number of active channels per patch (N) was significantly higher in KO mice, resulting in a higher channel activity (NP_O) in the absence of PON3. Although whole kidney abundance of ENaC subunits was not altered in Pon3 KOs, ENaC gamma subunit was more apically distributed within the connecting tubules (CNT) and cortical collecting ducts (CCDs) of Pon3 KO kidneys. Additionally, siRNA-mediated knockdown of PON3 in cultured mouse CCD cells led to an increased surface abundance of ENaC gamma subunit. As a result of lower plasma [K⁺], NCC phosphorylation was enhanced in the KO kidneys, a phenotype that was corrected by a high K⁺ diet. Finally, PON3 expression was upregulated in mouse kidneys under dietary K⁺ restriction, potentially providing a mechanism to dampen ENaC activity and associated K⁺ secretion. Taken together, our results show that PON3 has a role in renal Na⁺ and K⁺ homeostasis through regulating ENaC functional expression in the distal nephron.

Graphical Abstract

INTRODUCTION

The three mammalian paraoxonases (PON1, PON2 and PON3) share a high degree of similarity at sequence and structural levels, while exhibiting diverse substrate specificity (Mackness et al., 1996; Harel et al., 2004; Precourt et al., 2011). The founder of the family, PON1, hydrolyzes paraoxon and related organophosphates, whereas PON2 and PON3 only exhibit lactonase and lactonase/arylesterase activity (Oooms & Boter, 1965; Teiber et al., 2003; Draganov et al., 2005). However, all three PONs exhibit antioxidant properties and provide protection against lipid peroxidation (Mackness et al., 1991; Draganov et al., 2000; Ng et al., 2001). Altered serum paraoxonase activity and numerous PON1 polymorphisms have been associated with disorders of aging, including chronic inflammatory conditions, metabolic syndrome, liver impairment, kidney disease, cardiovascular disease, atherosclerosis, diabetes mellitus, neurodegenerative disorders, and cancer (Harel et al., 2004; Ng et al., 2005; Levy et al., 2019).

Several murine models have been developed to examine the physiological roles of mammalian PONs. Deletion of Pon1 in mice increases oxidative stress and accelerates atherosclerosis (Shih et al., 1998; Rozenberg et al., 2003). Similarly, Pon2 deficient mice are prone to develop diet-induced atherogenesis due to elevated mitochondrial oxidative stress and inflammation (Ng et al., 2006; Devarajan et al., 2011). Being the least studied member of the family, Pon3 KO mice are also more susceptible to obesity, gallstone formation, and atherosclerosis when challenged with high-fat diets (Shih et al., 2015). Mammalian PONs share structural homology with the C. elegans ortholog MEC-6, an ER chaperone that facilitates the functional expression of C. elegans degenerin channels and is thus required for worms’ response to gentle body touch (Huang & Chalfie, 1994; Chelur et al., 2002). We have previously shown that both PON2 and PON3 have roles in regulating the functional expression of the epithelial Na⁺ channel (ENaC), a member of the ENaC/degenerin family of ion channels, suggesting that there is functional conservation between mammalian PONs and their C. elegans orthologs (Shi et al., 2017; Shi et al., 2020; Shi et al., 2022).

ENaC is located in the later distal convoluted tubule (DCT2), connecting tubule (CNT), and collecting duct (CD) of ASDN (Meneton, 2000; Loffing et al., 2001). At the apical lumen of the ASDN, Na⁺ entry through ENaC provides the driving force for K⁺ secretion and thus has an essential role in regulating extracellular [K⁺], extracellular fluid volume, and blood pressure (BP) (Woda et al., 2001; Perrier et al., 2016; Boscardin et al., 2017). Patients with loss-of-function ENaC mutations are hyperkalemic with low BP (Geller et al., 1998; Adachi et al., 2001), whereas Liddle patients with gain-of-function ENaC mutations are hypertensive and hypokalemic with renal Na⁺ retention and volume expansion (Tamura et al., 1996; Uehara et al., 1998; Yamashita et al., 2001; Hiltunen et al., 2002). These changes are in part due to altered ENaC activity, however, there are also secondary effects of the changes in extracellular [K⁺], as hypo- or hyper-kalemia can impact the phosphorylation and activity of the thiazide sensitive sodium chloride co-transporter (NCC) (Hadchouel et al., 2010; Rengarajan et al., 2014; Grimm et al., 2017; Boscardin et al., 2018). Therefore, functional expression of ENaC in the kidney needs to be tightly regulated to maintain salt and volume homeostasis.

The role of PONs in kidney function and BP regulation remains obscure. Pon1 KO mice are hypotensive with hypoaldosteronism, whereas silencing Pon2 in rat kidneys results in hypertension (Gamliel-Lazarovich et al., 2012; Yang et al., 2012). We have previously shown that PON3 is primarily expressed in the ASDN (Shi et al., 2020), a segment with a prominent role in salt and water handling, and therefore BP control. However, its role in regulating these processes remains unknown. The present study was designed to investigated whether PON3 is involved in renal Na⁺ and K⁺ homeostasis by examining the functional expression of ENaC in a Pon3 KO murine model (Shih et al., 2015). Our results show that Pon3 KO mice have higher ENaC activity, elevated ENaC-mediated Na⁺ reabsorption, lower blood [K⁺], and elevated NCC phosphorylation. This is likely a result of upregulated ENaC functional expression in the distal nephron segments in the absence of PON3, demonstrating the importance of this protein for renal Na⁺ homeostasis.

METHODS

Ethical approval

All mice experiments were approved by the Institutional Animal Care and Use Committee of the University of Pittsburgh (protocol # 20128471). All procedures were performed following the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health. The authors have carefully reviewed and confirmed that the mice experiments performed in this paper comply with the animal ethics policies of The Journal of Physiology as detailed in Grundy (2015). Adult male and female mice were housed at the University of Pittsburgh animal facility on a 12-hour light/12-hour dark cycle with constant temperature and humidity control. A standard chow and water were provided ad libitum unless otherwise noted. Mice were sacrificed under isoflurane anesthesia with terminal cardiac puncture with heparinized 22-gauge needles. Care was taken to reduce the number of animals required for each experiment and to minimize any pain and distress of animals during any procedures.

Animals

The global Pon3 KO mouse line B6.129X1-Pon3^tm1Lus/J (PRID: IMSR_JAX: 027311) was originally developed by Dr. Aldons J. Lusis. Exon 4 of the Pon3 gene was replaced with a neomycin cassette, which results in a premature stop codon (Shih et al., 2015). Heterozygous male and female mice were crossbred to generate the experimental knockout mice (Pon3 KO) and wild type littermate controls (WT). Pups were genotyped in a single PCR reaction using the following primers: the common forward primer 5′-TGGGTATGTGAGGATCATTGC-3′, the WT reverse primer 5′-CCTTGACAGCCCTTCTCTGT-3′, and the mutant reserve primer 5′-GCCAGAGGCCACTTGTGTAG-3. The effects of altered dietary potassium were investigated using commercially available diets (Teklad), that were either low potassium that contained <0.003% K⁺ (TD.88239) or high KCl diet that contained 5% K⁺ (TD.09075).

Metabolic cage studies

Pon3 KO mice and WT controls were individually housed in wire-bottomed metabolic cages (Tecniplast). Mice were allowed to acclimate for at least 24 hours and were provided with gel diets made from chow with 1-2% agar along with drinking water. Baseline measures of water intake and urine output were taken over at least two 24-hour periods and data were averaged for each individual mouse. To assess amiloride sensitivity, animals were given a sham injection (10% DMSO in saline) on one night and an injection of amiloride (5 mg/kg) on a separate night, starting around 9 PM. Urine was collected at 3 h post-injection and analyzed for electrolyte content and urine volume. Urinary concentrations of Na⁺, K⁺, and Cl⁻ were measured using an EasyLyte analyzer (Medica). Urinary excretion of Na⁺ (U_NaV), K⁺ (U_KV), or Cl⁻ (U_ClV) was normalized to body weight (BW) and shown as nmol/min/gBW. Urine osmolality was measured with an Osmette Micro-Osmette Osmometer (Precision Systems).

Blood electrolytes and hormones

Pon3 KO mice and WT controls whole blood electrolytes were measured immediately at the point of blood draw using i-STAT Chem8+ cartridges (Abott #09P31-26). Plasma was separated via centrifugation at 2000 rpm for 7 min at 4 °C. Plasma aldosterone levels were measured using a commercially available ELISA kit according to the manufacturer's directions (Enzo Life Sciences #ADI-900-173).

Electrophysiology

Single channel activity was measured in freshly isolated mouse kidney tubules. Briefly, kidneys were surgically harvested from Pon3 KO mice and WT controls. The cortex regions were immediately separated from the whole kidney and minced into thin slices (~0.5 mm) and incubated with L-15 medium (Gibco #11415-064) containing 1mg/mL type II collagenase (Sigma #C6885) at 37°C for 25-45 minutes to loosen connecting tissues. Distal nephron segments likely containing DCT2, CNT and CCD were manually isolated under a microscope based on the unique bifurcation morphology. Isolated tubules were immobilized on cover glasses coated with Poly-L-lysine (Sigma #P4832) and transferred to a perfusion chamber. The single tubules were equilibrated in a bath solution (150 mM NaCl, 5 mM KCl, 1 mM CaCl₂, 2 mM MgCl₂ and 10 mM HEPES, pH 7.4) for at least 15 minutes before being manually split open near CNT/CCD segments, where the branches can be easily observed. Fire-polished micropipettes (3~6 mΩ) were filled with 140 mM LiCl, 2 mM MgCl₂ and 10 mM HEPES. Once a high resistance seal (GΩ) was obtained, single channel activity was measured in the cell-attached mode using a PC-One amplifier and a DigiData 1322A interface. Recordings were acquired at 5 kHz and filtered at 1 kHz with a built-in Bessel filter. To calculate single channel conductance, unitary channel activity was measured at different holding potentials to generate I-V curves. ENaC activity (NP_O) was assessed only in long recordings (> 5 min) of channels with characteristic ENaC kinetics (long mean open and closed times) at a holding potential of 60 mV. Electrophysiological data were analyzed with Clampfit 10.2 (Molecular Devices, RRID: SCR_011323).

Immunostaining

Formalin-fixed kidney tissues were embedded in paraffin and cut into 4 μm-thick sections by University of Pittsburgh Biospecimen Core. H&E staining was performed in four WT and four Pon3 KO mice. Images were taken on a Leica DM6000 widefield microscope. For immunofluorescence (IF) staining, kidney sections were deparaffinized, rehydrated, and blocked with 10% horse serum to block non-specific staining. Following an overnight incubation with rabbit anti-γENaC, goat anti-PON3, and rat anti-CaBP_28K antibodies at 4 °C, sections were washed and incubated with Jackson ImmunoResearch fluorescent secondary antibodies raised in donkey (anti-rabbit Cy3, anti-goat Alexa647 and anti-rat Alexa488) for two hours at room temperature. All antibodies were diluted in CytoVista^™ dilution buffer (ThermoFisher #V11305) as listed in Table 1. After extensive washing to remove excess secondary antibodies, sections were mounted in SlowFade antifade mountant (Thermofisher #S36917) and imaged on a Leica SP8 confocal microscope with a 40X oil objective. γENaC cellular distribution in the cortex segments was quantified using the line profile analysis of ImageJ software (RRID: SCR_003070). To do so, a line was drawn from the apical side to the basolateral side in both CalB_28K-positive DCT/CNT cells and CalB_28K-negative PCs of CCDs. γENaC staining intensity plotted along the distance from the apical side in individual cell was then aligned at the apical peak within each group. Three to four fields were imaged randomly for each kidney sample, and a total of three Pon3 KO and three WT kidneys were imaged.

Table 1. List of primary and secondary antibodies used in western blot (WB) or immunofluorescent (IF) staining.

The three newly developed BiCell antibodies were characterized in the present study based on their cell type specific localization and apparent molecular weight of proteins recognized by these antibodies.

Antibody	Source	Catalog #	PRID	Dilution (WB)	Dilution (IF)
goat anti- PON3	R&D Systems	AF4345	AB_2284023	1:2000	1:200
Rabbit anti-αENaC	Dr. Loffing lab	N/A	N/A	1:3000
Rabbit anti-βENaC	Stressmarq	SPC-404	AB_10644173	1:2000
Rabbit anti-γENaC	Stressmarq	SPC-405	AB_10640369	1:3000	1:200
Rabbit anti-NCC Thr53	Phospho Solutions	p1311-53	AB_2650477	1:1000
Rabbit anti-NCC	Millipore	AB3553	AB_571116	1:2000
Rat anti-ATP6V1B1	BiCell	20901	N/A	1:2000
Rabbit anti-Pendrin	BiCell	20501	N/A	1:1000
Rat anti-CaBP28K	BiCell	02401	N/A		1:500
Mouse anti-β-actin	Sigma	A1978	AB_476692	1:5000
Mouse anti-HA	Covance	MMS-101P	AB_2314672	1:3000
Mouse anti-FLAG	Sigma	F3165	AB_259529	1:3000
Donkey anti-goat HRP	Jackson ImmunoResearch	705-035-147	AB_2313587	1:5000
Donkey anti-mouse HRP	Jackson ImmunoResearch	715-035-151	AB_2340771	1:5000
Donkey anti-rabbit HRP	Jackson ImmunoResearch	711-035-152	AB_10015282	1:5000
Donkey anti-rat HRP	Jackson ImmunoResearch	712-035-153	AB_2340639	1:5000
Donkey anti-rabbit Cy3	Jackson ImmunoResearch	711-165-152	AB_2307443		1:1000
Donkey anti-goat Alexa647	Jackson ImmunoResearch	705-605-147	AB_2340437		1:500
Donkey anti-rat Alexa488	Jackson ImmunoResearch	712-545-150	AB_2340683		1:200

Immunoblotting

Flash-frozen kidney tissues were individually homogenized in Goldstein buffer (20 mM HEPES, 100 mM NaCl, 40 mM KCl, 1 mM EDTA, 10% glycerol, 1% NP40, 0.4% deoxycholate, pH 7.4) containing protease and phosphatase inhibitors (1:100, Sigma #539134, #P0044). Protein concentration was measured using a Pierce BCA protein assay kit (ThermoFisher #23227) and samples were denatured with 2X Laemmli sample buffer (Bio-Rad #1610737) at 37 °C for 30 minutes. 40 μg whole kidney lysate was subjected to SDS-PAGE separation under reducing conditions (5% β-mercaptoethanol, Bio-Rad #1610710) and transferred to a nitrocellulose membrane (Millipore #HAWP04700). The membranes were blocked with Tris-buffered saline (TBS) containing 5% skim milk for 1h at room temperature, then incubated with primary antibodies (listed in Table 1) diluted in 1% skim milk overnight at 4° C. After wash 3X with TBS, membranes were incubated with Jackson ImmunoResearch horseradish peroxidase conjugated secondary antibodies (Table 1, :5000, diluted in 1% milk) at room temperature for 1-2 hours. Blots were developed using Clarity Western ECL Substrates (Bio-Rad #1705060) and imaged with a Bio-Rad ChemiDoc^™. The intensity of the band of interest was quantified using Image Lab software (Bio-Rad, RRID:SCR_014210). For data normalization, we used either Bio-Rad stain-free technology to quantify the total protein loading, or the housekeeper protein β-actin.

ENaC surface biotinylation

immortalized mouse cortical collecting duct (mCCD, RRID: CVCL_B7HR) C1 cells were grown at 37 °C in DMEM/F12 medium ( Gibco #21041) supplemented with 5 μg/ml insulin (Sigma #I1882), 5 μg/ml transferrin (Sigma #T1428), 10 ng/ml epidermal growth factor (Sigma #E4127), 1 nM triiodothyronine (Sigma #T5516), 50 nM dexamethasone (Sigma #D8893), 0.06 nM sodium selenite (Sigma #S9133), and 2% decomplemented fetal bovine serum (FBS, Gibco #26140079) with 5% CO₂ as described previously (Klemens et al., 2017). Cells were reverse transfected with 100 picomoles of a control siRNA of scrambled sequence or a dicer specific siRNA targeting the mouse Pon3 sequence 5’-GUACUAUAUUUCACAAAGCUCUGTA-3’ (Integrated DNA Technologies) on 24mm polyester membrane inserts (Corning #3450) using Lipofectamine 3000 (Invitrogen #L3000008) according to the manufacturer’s instruction. Transfected cells were cultured for an additional 3-4 days to fully polarize. Apical surface ENaC were labeled with 1 mg/mL of EZ-Link^™ Sulfo-NHS-SS-Biotin (ThermoFisher #21331) in a buffer containing 137 mM NaCl, 15 mM sodium borate, pH adjusted to 9.0. Excess biotin was then quenched with 10% FBS in DMEM/F12. After extensive washes with PBS, cells were extracted in Goldstein buffer. 5% of the total lysate was retained as “whole cell” while the remainder was incubated with Streptavidin agarose (ThermoFisher #20349) overnight at 4°C to isolate biotinylated “surface” proteins. The recovered surface proteins, along with the 5% of the whole cell lysate, were separated by SDS-PAGE to probe for γENaC in the top half of the blot, and for PON3 or β-actin in the bottom half of the blot.

Co-immunoprecipitation (CO-IP)

Fisher Rat Thyroid (FRT, RRID: CVCL_A6IE) cells were routinely cultured in DMEM/F-12 medium supplemented with 9% FBS. Cells were seeded in a 6-well dish and transfected with 0.5 μg of plasmids encoding mouse NCC with a N-terminal HA epitope tag (HA-NCC), mouse PON3 with a C-terminal FLAG epitope tag (PON3-FLAG, OriGene #MR220409), or both plasmids (NCC + PON3) using Lipofectamine 3000. 24h post-transfection, cells were extracted in Goldstein buffer supplemented with protease inhibitor. 10% of the total lysate was saved as “Input”. The remaining was incubated with either 1μg anti-HA antibodies to pulldown NCC or with 1μg anti-FLAG antibodies to pull down PON3. The precipitation was achieved by end-over-end mixing overnight at 4°C in the presence of 50 μL rec-protein G-Sepharose (Invitrogen #10-1241). Proteins associated with the beads were eluted into 30 μL 2X Laemmli sample buffer containing 5% β-mercaptoethanol. Both the IP and the input were subjected to SDS-PAGE and blotted either for PON3 or for NCC. Assays were repeated three times for each condition.

Statistical analyses

Data were expressed as the mean ± S.D. in the main text and shown in Scatter-dot plots with a horizontal bar indicating the mean. Data distribution was examined with a D'Agostino-Pearson normality test. Statistical comparisons between two groups were determined with one-sample Wilcoxon test or nonparametric Mann-Whitney test. Statistical comparisons between three or more groups were determined with one-way or two-way ANOVA followed by a Šidák’s multiple comparisons test, using GraphPad Prism Software (RRID:SCR_002798). A p value of < 0.05 was considered statistically significant. Analysis method and exact p values are clearly stated in all figures and tables.

Data availability

All original blots and microscopic images have been provided to the journal. All original electrophysiological recordings used to generate data presented in figure 3 are archived and are fully available to the journal upon reasonable request.

Figure 3. ENaC channel number (N) and activity (NP_O) are higher in Pon3 KO mice.

Distal nephron segments (including DCT, CNT and CD) were manually dissected based on the unique bifurcation morphology. Isolated tubules were manually split open to expose the apical side of PCs within the CNT/CD segments. ENaC single channel activity was recorded using cell-attached mode with Li⁺ as the conducting ion. (A) Representative gap-free current traces at different holding potentials, ranging from +20 to +100 mV, recorded in tubules isolated from Pon3 KO mice or WT littermates; (B) Summary plot of the I-V relationship of ENaC recording in Pon3 KO mice and WT littermates; (C) Representative gap-free current traces used for NP_O estimation, with a pipette potential of +60 mV; (D-F) Summary data of NP_O, apparent N and estimated P_O of the Pon3 KO mice and their WT littermates (females in red and males in blue). Statistical comparison was made with Unpaired Student’s t-test with Welch’s correction and p values are shown. (G) Summary of empty patches vs patches containing ENaC currents in five Pon3 KO mice and seven WT littermates. Both sexes are included.

RESULTS

ENaC-mediated Na⁺ reabsorption is increased in Pon3 KO mice at baseline

We obtained a global Pon3 KO mouse, in which exon 4 of the Pon3 gene was replaced with a neomycin cassette (Shih et al., 2015) (Fig. 1A). The genotypes of experimental mice were confirmed by PCR using DNA isolated from tail snips (Fig. 1B) and PON3 knockout was further confirmed by immunoblotting of whole kidney lysates (Fig. 1C). At baseline, Pon3 KO mice had normal kidney histology without apparent inflammation or injury (Fig. 1D). To test whether deletion of PON3 altered renal Na⁺ and K⁺ handling in mice, we collected 24 h urine samples from Pon3 KO mice and WT littermates individually housed in metabolic cages while body weight (BW), water consumption, and urine output were recorded daily. There were no apparent differences in daily water consumption, urine volume, or urine osmolality between Pon3 KO mice and WT littermates (Table 2). Urinary excretion of Na⁺ (U_NaV), K⁺ (U_KV), and Cl⁻ (U_ClV) over 24h were also similar between the two groups (Table 2). When compared to their WT littermates, Pon3 KO mice had a lower blood [K⁺], largely due to changes in the KO females (Table 3). No significant differences were observed for plasma [Na⁺], tCO2 or aldosterone (Aldo) levels between the two groups of mice for either sex (Table 3).

Figure 1. Global Pon3 KO mice have normal kidney morphology.

(A) Pon3 KO mice were generated by replacing exon 4 with a neomycin cassette (Neo) via homologous recombination. Both the forward primer (p1) and reverse primer (p2) are located within intron 3 and can amplify the WT allele. The reverse primer p3 was designed to amplify the Neo cassette in the KO allele when paired with p1. (B) Representative genotyping results of tail samples from the WT (+/+), KO (−/−), or heterozygotes (+/−) Pon3 mouse. The DNA ladder is shown at the left to indicate the sizes of PCR products. (C) Western blot analysis of whole kidney lysates from three WT and three Pon3 KO mice, probed for PON3 and the loading control Actin. Protein ladder is shown at the left to indicate the molecular weight of the target bands. (D) H&E staining of four WT (+/+) and four Pon3 KO (−/−) adult male kidneys. Three enlarged views are shown for the cortex (1), outer medulla (2), or inner medulla (3), with scale bars = 20 μm.

Table 2. Baseline urine parameters of mice maintained on regular chow.

Data shown as mean ± SD. The number of animals per group (n) and p values are shown. Statistical comparison between the Pon3 KO mice and their WT littermates was analyzed using one-way ANOVA followed by Šídák's multiple comparisons test.

	WT (F) (n = 8-13)	KO (F) (n = 7-8)	P value	WT (M) (n = 6-8)	KO (M) (n = 7-8)	P value	WT (F + M) (n = 16-19)	KO (F + M) (n = 14-16)	P value
Body weight (BW), g	21.7 ± 3.0	21.4 ± 2.9	0.997	28.9 ± 2.7	28.1 ± 2.6	0.960	24.4 ± 4.6	24.8 ± 4.3	0.990
24h water intake, mL	2.07 ± 0.79	2.08 ± 0.74	0.991	1.60 ± 0.85	1.71 ± 0.85	0.991	1.92 ± 0.81	1.90 ± 0.79	0.998
24h urine output, mL	1.38 ± 0.66	1.41 ± 1.00	0.999	2.38 ± 0.51	2.50 ± 0.76	0.994	1.70 ± 0.77	1.95 ± 1.03	0.745
24h UNaV, μmol/gBW	2.59 ± 0.94	2.68 ± 1.87	0.999	3.29 ± 0.55	3.84 ± 1.27	0.814	2.81 ± 0.88	3.26 ± 1.66	0.656
24h UKV, μmol/gBW	13.9 ± 4.1	13.9 ± 5.8	0.997	16.1 ± 2.7	15.7 ± 3.9	0.991	14.6 ± 3.7	14.8 ± 4.9	0.999
24h UClV, μmol/gBW	7.94 ± 2.57	7.89 ± 3.73	0.994	9.37 ± 1.56	9.62 ± 2.71	0.998	8.39 ± 2.36	8.76 ± 3.27	0.971
Urine osmolarity, Osm/kg	1.34 ± 0.23	1.20 ± 0.36	0.692	1.15 ± 0.13	1.06 ± 0.10	0.795	1.25 ± 0.21	1.10 ± 0.20	0.232

Table 3. Blood parameters of mice at baseline (normal K⁺ diet, NK) or receiving a high K⁺ diet (HK, 5% KCl) for 9 days.

Data shown as Mean ± SD. The number of animals per group (n) and p values are shown for both sexes. Statistical comparison between the Pon3 KO mice and their WT littermates on the same diets or between different dietary challenges was analyzed using two-way ANOVA followed by Šídák's multiple comparisons test.

Females		Normal K+ diet (NK)			High K+ diet (HK)
		WT (8-13)	KO (7-8)	p (vs WT)	WT (7-8)	p (vs NK WT)	KO (9-10)	p (vs NK KO)	p (vs HK WT)
	Plasma Na+, mEq/L	143.9 ± 1.4	144.9 ± 1.2	0.333	143.4 ± 1.8	0.658	142.5 ± 2.6	0.019	0.336
	Plasma K+, mEq/L	4.62 ± 0.24	4.24 ± 0.37	0.050	4.86 ± 0.49	0.223	4.56 ± 0.32	0.085	0.105
	Plasma Aldo, pg/mL	308 ± 206	365 ± 304	0.706	1237 ± 364	<0.001	933 ± 320	<0.001	0.047
	Plasma tCO2, mEq/L	23.3 ± 1.4	23.4 ± 2.0	0.260	21.7 ± 2.7	0.293	21.0 ± 1.9	0.398	0.282
Males		Normal K+ diet (NK)			High K+ diet (HK)
		WT (6-8)	KO (7-8)	p (vs WT)	WT (10)	p (vs NK WT)	KO (6-7)	p (vs NK KO)	p (vs HK WT)
	Plasma Na+, mEq/L	143.4 ± 1.6	145. 3 ± 1.1	0.142	144.4 ± 2.3	0.383	141. 8 ± 4.2	0.017	0.051
	Plasma K+, mEq/L	4.56 ± 0.35	4.36 ± 0.28	0.375	4.85 ± 0.54	0.179	4.90 ± 0.50	0.035	0.827
	Plasma Aldo, pg/mL	249 ± 171	310 ± 135	0.813	1547 ± 843	<0.001	991 ± 487	0.022	0.049
	Plasma tCO2, mEq/L	23.0 ± 2.0	24.3 ± 1.8	0.865	21.9 ± 2.1	0.150	21.3 ± 5.3	0.020	0.476
Females + Males		Normal K+ diet (NK)			High K+ diet (HK)
		WT (16-19)	KO (14-16)	p (vs WT)	WT (17-18)	p (vs NK WT)	KO (15-17)	p (vs NK KO)	p (vs HK WT)
	Plasma Na+, mEq/L	143.6 ± 1.4	145.1 ± 1.1	0.059	144.0 ± 2.1	0.601	142.3 ± 3.2	0.003	0.085
	Plasma K+, mEq/L	4.60 ± 0.30	4.30 ± 0.32	0.048	4.85 ± 0.51	0.067	4.69 ± 0.42	0.010	0.238
	Plasma Aldo, pg/mL	278 ± 186	335 ± 222	0.698	1410 ± 675	<0.001	957 ± 380	<0.001	0.003
	Plasma tCO2, mEq/L	23.1 ± 1.7	23.9 ± 1.9	0.342	21.8 ± 2.3	0.079	21.1 ± 3.4	0.008	0.903

Amiloride and its derivatives exert their potassium-sparing and natriuretic effects by inhibiting ENaC activity in the distal nephron (Bull & Laragh, 1968; George et al., 1989; Sun & Sever, 2020). We performed amiloride-induced natriuresis to determine whether the lower blood [K⁺] may reflect, in part, an upregulation in ENaC-dependent Na⁺ reabsorption and K⁺ secretion in Pon3 KO mice. Mice were maintained on standard chow and given a single intraperitoneal injection of saline (10% DMSO, vehicle control) or amiloride (5mg/kg BW). Urinary output, U_NaV, U_KV, and U_ClV were compared between Pon3 KO mice and WT littermates at 3 hours post-injection (Fig. 2A-D). As expected, amiloride significantly increased U_NaV in both WT mice (5.5 ± 2.0-fold, n=13, p < 0.01) and Pon3 KO mice (6.7 ± 0.5-fold, n=10, p < 0.01) (Fig. 2B). The reduction in U_KV did not reach statistical significance in either group (Fig. 2C). U_ClV was also significantly increased by amiloride injection in both groups (Fig. 2D). Most importantly, amiloride-induced change in U_NaV (ΔU_NaV) was significantly higher in Pon3 KO mice when combining mice of both sexes (n=10, p = 0.048) (Fig. 2F), suggesting that ENaC-dependent Na⁺ reabsorption was increased in Pon3 KO mice, a likely reason for the minor hypokalemic phenotype observed in the Pon3 KOs (Table 3).

Figure 2. Amiloride-induced natriuresis is enhanced in Pon3 KO mice.

Mice were given a single IP injection of saline (10% DMSO, vehicle control) or 5 mg/kg amiloride and urine samples were collected 3h post-injection. (A) 3h urine volume of individual mice after saline or amiloride injection. (B-D) Urinary excretion of Na⁺ (U_KV), K⁺ (U_KV) or Cl⁻ (U_ClV) was calculated per gram of body weight (gBW) and expressed as the rate of excretion per minute. (E-H) Amiloride-induced changes (Δ) in urine volume, U_NaV, U_KV or U_ClV of females (in red) or male (in blue) Pon3 KO mice and their WT littermates are shown. A total of six WT males, seven KO males, seven WT females, and three KO females were used in two separate metablic cage studies. Statistical comparisons between the groups were analyzed using two-way ANOVA followed by Šídák's multiple comparisons test. p values are shown for all comparisons.

ENaC activity is enhanced in the CNT/CCD segments of Pon3 KO mice

To directly measure endogenous ENaC activity, we performed single channel recording in split-open tubules of Pon3 KO mice or their WT littermates. To gain access to the apical side of principal cells (PCs), freshly isolated tubules were manually split open and ENaC activity was recorded in the cell-attached mode with Li⁺ as the main conducting ion. As shown in figure 3A-B, the average single channel conductance was 7.8 ± 2.2 pS in Pon3 KO mice (n=10), similar to the conductance in their WT littermates (7.4 ± 2.7 pS, n=8, p = 0.69). ENaC activity (NP_O) was assessed in long recordings (> 5 min) from channels with characteristic ENaC kinetics (long mean open and closed times, shown in Fig. 3C). The number of active channels per patch (N) and channel open probability (P_O) was estimated as previously described (Shi et al., 2016). As shown in figure 3D, Pon3 KO mice had higher NP_O than their WT littermates. The estimated channel P_O was similar between the two groups of animals (Fig. 3E), whereas channel number (N) was significantly higher in Pon3 KO mice (Fig. 3F). Together, our results suggest that the higher ENaC activity in Pon3 KO mice is mainly due to more active channels at the apical side of PCs of CNT/CCD segments. In addition, 19 out of a total of 35 patches across seven WT mice displayed no ENaC activity, whereas only six out of 24 patches lacked ENaC activity in a total of five Pon3 KO mice (Fig. 3G). The apparently higher frequency of detecting ENaC-like activity in the KO kidneys provides additional evidence that channel abundance is enhanced in the absence of PON3.

γENaC cellular distribution in PCs of CNT/CCD is altered in Pon3 KO mice

Functional ENaC complexes consist of three homologous subunits, α, β and γ (Jasti et al., 2007; Stockand et al., 2008; Kashlan & Kleyman, 2011). We performed quantitative RT-PCR to examine transcript abundance of the three ENaC subunits. As shown in figure 4A, kidney abundance of Scnn1a, Scnn1b or Scnn1g transcripts was not altered in Pon3 KO mice maintained on standard chow, suggesting that deleting PON3 does not affect ENaC expression at the transcriptional level. Next, we performed immunoblotting analysis to determine whether the protein abundance of ENaC subunits is altered in Pon3 KO kidneys. As shown in figure 4B-C, Pon3 KO mice have similar whole kidney levels of αENaC (full length and cleaved form), βENaC, and γENaC (full length and cleaved form) when compared to their WT littermates. This is not surprising, as ENaC activity is largely determined by its surface expression rather than the total protein abundance. To determine ENaC cellular distribution, we performed immunofluorescent (IF) staining of γENaC in Pon3 KO mice or their WT littermates. Kidney sections were co-stained for Calbindin-D28K (CalB_28K), a marker of DCT/CNT, to demarcate the cortex region (Roth et al., 1982; Bauchet et al., 2011; Lee et al., 2015). As shown in Fig. 5A, γENaC signal was detected in tubules with strong CalB_28K staining, likely DCT2, as well as tubules with lower CalB_28K signal, likely CNT, and CCD within the cortex. More importantly, PON3 co-localized with γENaC in both DCT2 cells and PCs of CNT/CCD. To determine if the level of ENaC located at the apical surface was altered in Pon3 KO kidneys, γENaC intensity was quantified across a line-scan in cells with well-defined luminal vs basolateral sides. As shown in Fig. 5B, γENaC staining within DCT2 (tubules with high CalB_28K signal) was similar between Pon3 KOs and WT littermates. In contrast, the KO mice had significantly higher apical staining of γENaC in tubules showing reduced CalB_28K staining. Together, our data suggest that PON3 KO promotes ENaC surface trafficking, specifically in PCs of CNTs and CCDs.

Figure 4. Whole kidney abundance of the three ENaC subunits was not altered in Pon3 KO mice.

(A) Quantitative RT-PCR analysis of transcript levels of the ENaC subunits. Total RNA was isolated from kidney cortex of Pon3 KO mice or their WT littermates and used as templates for qPCR analysis. The message levels of each ENaC subunit were normalized to that of 18S levels in the same sample. Five WT or KO male mice and four WT or KO female mice were analyzed. (B) Representative blots showing αENaC (full length and cleaved), βENaC, and γENaC (full-length and cleaved) in kidney lysates of male and female Pon3 KO mice or their WT littermates. Images of stain-free gels are shown at the bottom for total protein estimation. Markers of molecular weight (kDa) are shown on the left. (C) The abundance of each ENaC subunit was normalized to total protein in the corresponding sample and are shown as relative values of the WT mice of the same sex. Statistical comparison between the WT and Pon3 KO mice was analyzed using two-way ANOVA, followed by Šídák's multiple comparisons test and p values are shown. Each point represents an individual animal, with a total of nine WT males, six KO males, nine WT females and seven KO females being utilized for two separate blots.

Figure 5. γENaC subcellular distribution is more luminal in Pon3 KO kidneys.

(A) 4 μm kidney sections of Pon3 KO mice and their WT littermates were co-stained for γENaC (red), Calbindin 28kD (CalB_28K, green) and PON3 (blue), scale bars = 30 μm. Maximum intensity projections of z-stacks are shown to visualize γENaC intensity and localization. Specific nephron segments are annotated in the overlay images. Zoom-in views of γENaC staining in the CalB_28K-postive DCT2 cells (B) and CalB_28K-negative CNT/CCD cells (C) are shown, scale bars = 5 μm. γENaC intensity was quantified using the Intensity Plot Profile Function in ImageJ with a line drawn from the apical lumen to the basolateral side, as shown with dashed arrows. The average intensity and standard errors are plotted along the distance from the apical side. The number of representative cells from WT mice and Pon3 KO mice are noted in the corresponding panel.

γENaC surface expression is enhanced by PON3 knock down (KD) in mCCD cells

To directly examine the effect of PON3 on ENaC surface expression, we utilized an immortalized mouse cortical collecting duct cell line (mCCD). Previously, we had shown that Pon3 knock down (KD) increased amiloride-sensitive Na⁺ transport in mCCD cells (Shi et al., 2020). We performed biotin labeling of mCCD cells transiently transfected with Pon3 specific siRNA (PON3 KD) or with scrambled siRNAs (negative control, NC). Biotinylated surface proteins were isolated with Streptavidin and blotted for γENaC in both the whole cell lysates and biotinylated surface fractions (Fig. 6A). Actin was only detected in the whole cell lysate input, suggesting that the surface pulldown was free of cytosolic contamination. When PON3 expression was reduced by 42 ± 11% in the whole cell (n=6 for each group, p = 0.02 KD vs. NC), surface expression of full-length γENaC was increased by 65 ± 22% (p < 0.01 KD vs. NC) and the cleaved form was increased by 75 ± 21% (p < 0.01 KD vs. NC) (Fig. 6B). Consistent with our previously published observation, whole cell abundance of γENaC was also increased by PON3 KD in cells (52 ± 26% and 84 ± 39% for the full length and cleaved form, respectively (p < 0.01 KD vs. NC)). These data suggest that PON3 negatively regulates the functional expression of endogenous ENaC in mCCD cells.

Figure 6. Pon3 knockdown (KD) enhances γENaC surface expression in mCCD cells.

mCCD cells were transiently transfected with Pon3 specific siRNA (PON3 KD) or with scrambled siRNAs (negative control, NC) and cultured on permeable inserts to form a confluent monolayer. Apical surface ENaC were labeled with impermeable Sulfo-NHS-SS-biotin and recovered using Streptavidin beads. (A) The top half of the blot was incubated with anti-γENaC antibody to probe ENaC abundance in the biotinylated surface fractions compared with the 5% total whole cell lysates (top panel). The bottom half of the blot was incubated with anti-PON3 antibody to confirm Pon3 KD efficiency (middle panel) and stripped to re-probe for actin for both loading control and surface-pulldown control (bottom panel). Markers of molecular weight (kDa) are shown on the left; (B) Relative surface and total abundance of γENaC (full length and cleaved) as well as whole cell PON3 were estimated and shown as relative values to the NC cells. Statistical comparison between the Pon3 KD cells and NC cells was analyzed using two-way ANOVA, followed by Šídák's multiple comparisons test and p values are shown.

NCC phosphorylation is upregulated in Pon3 KO mice

Several groups have reported a negative relationship between plasma [K⁺] and NCC phosphorylation (Terker et al., 2016; Boyd-Shiwarski et al., 2020). As Pon3 KO mice have lower blood [K⁺] (see Table 3), we hypothesized that NCC phosphorylation would be higher in this group as well. We performed immunoblotting to detect total NCC (tNCC) and phosphorylated NCC (T⁵³-NCC, pNCC) in kidney lysates of Pon3 KO mice and their WT littermate controls (Fig. 7A). We found that pNCC abundance was increased 76 ± 85% in male KO mice (n=8, p < 0.01) and 55 ± 45% in female KO mice (n=7, p < 0.01), whereas tNCC abundance was not altered in the KO animals when compared to their WT littermates (Fig. 7B). In addition, we did not observe PON3-NCC protein complexes through co-immunoprecipitation, nor were any regulatory effects of PON3 on NCC expression in HEK cells observed (Fig. 8). Together, our data suggest that the increase in NCC phosphorylation in the kidney is likely due to chronic lower blood [K⁺] in Pon3 KO mice and not a direct role of PON3 in regulating NCC expression.

Figure 7. NCC phosphorylation is upregulated in Pon3 KO mice, but total expression is not altered.

(A) Kidney lysates of Pon3 KO mice and their littermates were incubated with antibody against pT⁵³-NCC (pNCC) and subsequently stripped and re-probed for total NCC (tNCC). The bottom half of the blot was used for detecting PON3. Images of stain-free gels are shown at the bottom for total loaded protein estimation. Markers of molecular weight (kDa) are shown on the left; (B) The abundance of pNCC and NCC was normalized to the total protein in the corresponding sample and are shown as relative values of the WT mice of the same sex. Statistical comparison between the Pon3 KO mice and their WT littermates was analyzed using two-way ANOVA, followed by Šídák's multiple comparisons test and p values are shown. Each point represents an individual animal, with 8 KO males, 8 WT males, 7 KO females and 9 WT females being analyzed across two separate blots.

Figure 8. PON3 does not interact with NCC or affect its expression.

FRT cells were transiently transfected with plasmids encoding HA-NCC, PON3-FLAG, or both constructs (NCC + PON3). 10% of the lysates were saved as “input” and the remainder was incubated either with anti-HA antibodies to pulldown NCC (A) or with anti-FLAG antibodies to pull down PON3 (B). The presence of PON3 or NCC in the IPs and whole cell lysates were examined, respectively. Markers of molecular weight (kDa) are shown on the left and right. The nonspecific band above 50 kDa in all three samples in the middle blot of panel (A) was denoted with *. (C) The effect of PON3 on NCC expression is shown. Statistical comparison was analyzed using one-sample Wilcoxon test with the p value shown.

Effects of Pon3 KO are rescued with a high K⁺ diet

To determine whether Pon3 KO mice have compromised adaptation to dietary K⁺ challenge, adult Pon3 KO mice and their littermates were fed with a high K⁺ (HK) diet (5% K⁺ as KCl) for 9 days. While a HK diet significantly increased plasma Aldo levels in both groups when compared to mice maintained on a normal K⁺ (NK) diet, Pon3 KO mice had significantly lower plasma Aldo levels than their WT littermates on the HK (Table 3). Importantly, significant differences in plasma [K⁺] that existed with NK diet were no longer present between Pon3 KO mice and their WT littermates when fed with a HK diet (Table 3). When plasma [K⁺] in Pon3 KO mice was normalized by the HK diet, levels of NCC phosphorylation between the two groups were no longer different (Fig. 9 A-B). In addition, γENaC abundance in Pon3 KO mice was similar to their littermate controls (Fig. 9 C-D). While these data suggest that dietary K⁺ supplementation largely overrides the effect of deleting Pon3 on ASDN Na⁺/K⁺ handling, the significantly lower Aldo levels observed in the Pon3 KO mice on the HK diet, compared to WT, suggests that less Aldo is required to adapt to the HK diet in Pon3 KO mice.

Figure 9. NCC phosphorylation in Pon3 KO mice was normalized by dietary K⁺ supplementation.

Male and female Pon3 KO mice and their WT littermates were fed with 5% K⁺ (as KCl) diet for 9 days. Kidney lysates were incubated with antibodies against pT⁵³-NCC (pNCC) (A) or against γENaC (C). Images of stain-free gels are shown at the bottom for total loaded protein estimation. Markers of molecular weight (kDa) are shown on the left. The abundance of pNCC (B) and γENaC (D) was normalized to total protein in the corresponding sample and are shown as relative values compared to WT mice of the same sex. Statistical comparison between the Pon3 KO mice and their WT littermates was analyzed using two-way ANOVA, followed by Šídák's multiple comparisons test, with p values shown. A total of five KO males, six WT males, six KO females, and six WT females were analyzed.

PON3 expression in mouse kidney is regulated by dietary K⁺

As deleting Pon3 in mice led to higher functional ENaC expression (Figs. 2 and 3) and lower plasma [K⁺] (Table 3), we hypothesized that PON3 abundance would change in response to conditions that influence Na⁺ and K⁺ homeostasis. To test this hypothesis, adult male C57BL/6J mice were fed with either a HK (5% K⁺) or low K⁺ (LK, <0.003% K⁺) diet for 3 or 10 days. Kidneys were harvested and examined for protein expression levels of PON3 and tNCC (Fig. 10A). Consistent with previous observations (Vallon et al., 2009; Frindt & Palmer, 2010; Kortenoeven et al., 2021), tNCC abundance was suppressed by the HK diet and upregulated with the LK diet (Fig. 10C). We found that PON3 protein abundance significantly increased when mice were fed the LK diet for 10 days (Fig. 10B). It has been shown that both ENaC abundance and its proteolytic cleavage are reduced in hypokalemic rodents (Elkjaer et al., 2002; Ayasse et al., 2021). The increase in PON3 abundance is a potential mechanism to reduce functional ENaC expression under dietary K⁺ restriction. Interestingly, PON3 expression was not altered in mice fed with the LK diet for a shorter period (3 days), suggesting that changes in PON3 expression in response to K⁺ restriction is a time dependent process. While we did not observe a significant change in PON3 expression in kidneys from mice fed with the HK diet for 10 days, the abundance of PON3 was similar to that in mice fed with the LK diet for 10 days (Fig. 10B).

Figure 10. The effect of K⁺ diets on renal PON3 expression.

Male C57Bl/6J adult mice were fed with either a low K⁺ diet (<0.003%) or a high K⁺ diet (5% KCl) for 3 or 10 days (LK3, LK10, HK3, HK10). Mice maintained on regular chow (Ctrl, ~1% K⁺) were included as controls. (A) Kidney lysates were incubated with antibody against PON3, tNCC, or β-actin as a loading control. Markers of molecular weight (kDa) are shown on the left. (B) The abundance of PON3 or tNCC was normalized to β-actin in the corresponding sample and are shown as relative values compared to Ctrl mice. Four mice per group were included for analysis. Statistical comparison between mice fed with different K⁺ diets and the ctrl mice was analyzed using one-way ANOVA, followed by Dunnett's multiple comparisons test, with p values shown.

Pon3 KO does not alter the expression of V-ATPase or Pendrin in Intercalated Cells

Single cell RNAseq of major renal cell types has revealed that Pon3 transcript is more abundant in intercalated cells (ICs) than PCs in murine kidneys (Chen et al., 2017; Lee et al., 2018). Indeed, we have previously detected high levels of PON3 protein in AQP2-negative ICs within the renal medulla via immunofluorescent staining (Shi et al., 2020). In fact, PON3 co-localizes with V-ATPase B1 in type A or non-A/non-B ICs, or with pendrin in type B ICs of the ASDN (Shi et al., 2020). These findings raise the question of whether PON3 could serve a chaperone function in ICs and potentially contribute to acid/base homeostasis. To explore the potential IC targets of PON3, we examined expression levels of pendrin and V-ATPase B1 subunit in kidneys of Pon3 KO mice and their littermate controls, as well as whole blood total CO₂ (tCO₂) levels. Deleting Pon3 did not alter the protein abundance of the V-ATPase B1 subunit in whole kidney lysate (Fig. 11 A-B) nor did it significantly alter whole blood tCO₂ levels (Table 3). In addition, pendrin protein expression was similar in kidney lysates from Pon3 KO and WT mice (Fig. 11 C-D). Our results suggest that PON3 is not required for pH regulation under baseline conditions and that neither pendrin nor the V-ATPase B1 subunit is a substrate of PON3’s chaperone function.

Figure 11. Expression of Pendrin and V-ATPase B1 subunit are not altered in kidneys from Pon3 KO mice.

Kidney lysates from Pon3 KO mice and their littermate controls were probed for V-ATPase B1 (A) or Pendrin (C). Images of stain-free gels are shown at the bottom for total loaded protein estimation and normalization. Markers of molecular weight (kDa) are shown on the left; the abundance of V-ATPase B1 subunit (B) or Pendrin (D) was normalized to total protein in the corresponding sample and are shown as relative values compared to WT mice of the same sex. Statistical comparison between the Pon3 KO mice and their WT littermates was analyzed using two-way ANOVA, followed by Šídák's multiple comparisons test, with p values shown. A total of six KO males, seven WT males, seven KO females, and eight WT females were analyzed across two separate blots.

DISCUSSION

While mammalian PONs are mainly recognized for their antioxidant properties, our work suggests that these proteins also function as molecular chaperones. We have previously demonstrated that PON2, which predominantly resides in the ER, promotes ENaC ubiquitination and proteasomal degradation (Shi et al., 2022), and this function of PON2 on ENaC expression is independent of its enzymatic activity (Shi et al., 2017). Additionally, we have shown that PON3 negatively regulates ENaC expression in vitro. Given that PON3 is primarily expressed within ASDN in mouse kidney (Shi et al., 2020) and its expression is altered by dietary K⁺ (Fig. 10), we hypothesized that PON3 regulates ENaC activity in vivo, as ENaC expression is highest in these segments within the cortex (Yang et al., 2021). If this hypothesis were true, we would expect the KO mice to have higher ENaC-mediated Na⁺ reabsorption and K⁺ secretion, given that ENaC expression at the membrane should increase. Indeed, Pon3 KO mice have lower plasma [K⁺] (Table 3) and higher amiloride-induced natriuresis (Fig. 2), suggesting an upregulated ENaC-mediated Na⁺ reabsorption. Additionally, our patch clamp experiments on split-open tubules directly show that Pon3 KO mice have higher ENaC activity (NP_O) which was mainly driven by a higher number of active channel (N) in the CNT/CCD segments (Fig. 3). Interestingly, we found that the cellular distribution of γENaC in Pon3 KO kidneys was shifted toward the luminal side in CNT/CCD (low CalB_28K-signal in Fig. 5), suggesting that PON3 affects ENaC surface trafficking specifically in PCs of CCDs. Consistent with our data in mice, there was increased apical surface expression of the cleaved and full length γENaC when the endogenous PON3 was knocked down by siRNA (Fig. 6). Together, our data suggest that PON3 regulates ENaC function within the distal nephron through modulating channel subunit surface expression. Upon challenge with a high K⁺ diet, Pon3 KO mice have normalized plasma [K⁺] (Table 3), as well as similar levels of NCC phosphorylation (Fig. 9) when compared to their littermate controls, suggesting that PON3’s effect on distal nephron Na⁺ and K⁺ transport is ablated by increased aldosterone levels and dietary K⁺. Indeed, PON3 expression in the kidney was significantly increased by dietary K⁺ restriction. Surprisingly, similar levels of kidney PON3 expression were seen in mice fed either a LK or HK diet (Fig. 10). While PON3 upregulation by a LK diet may function to suppress ENaC activity and prevent K⁺ wasting, the mechanism for a similar trend under the HK diet seems puzzling. One possibility is that PON3 expression within different nephron segments and/or cell types responds differently to changes in dietary K⁺. One caveat is that long-term treatment with extreme K⁺ diets (<0.003% for 14 days or 5% KCl for 10 days) could induce renal cell hypertrophy and hyperplasia which ultimately changes the morphology and structure of mouse kidneys (Stetson et al., 1980; Stanton et al., 1981; Fervenza & Rabkin, 2002). Changes in the size and/or number of PON3-positive cells relative to other cells in the kidney could confound our assessment of PON3 expression at the whole kidney level. Additional studies are needed to further investigate how dietary K⁺ regulates PON3 expression in specific cell types.

Many chaperones, including PONs have broad substrate specificities. In a murine Adriamycin-induced nephropathy model, Pon2 deficiency was shown to affect Ca²⁺ entry through the transient receptor potential cation channel subfamily C member 6 (TRPC6) which led to podocyte injury and aggravated proteinuria (Hagmann et al., 2022). Although it was proposed that PON2 affected TRPC6 activity by altering cell membrane lipid composition, other potential mechanisms of regulating channel activity including TRPC6 abundance, trafficking, or stability have not been examined (Hagmann et al., 2022). PON2 can stimulate glucose uptake by releasing the glucose-transporter (GLUT1) from its inhibitor stomatin and thus facilitates the growth and metastasis of pancreatic ductal adenocarcinoma (Nagarajan et al., 2017). A similar mechanism was employed by PON2 to promote B cell leukemogenesis, again demonstrating its role as a chaperone (Pan et al., 2021). As PON3 co-localizes with NCC in DCT2 (Shi et al., 2020), it is possible that NCC is a direct target of PON3’s chaperone function. However, we failed to detect any interaction between NCC and PON3 by co-immunoprecipitation (Fig. 8), and we found that neither PON2 (Shi et al., 2022) nor PON3 regulated the expression of NCC in an over-expression system (Fig. 8). In addition, only the levels of phosphorylated NCC, not the level of total NCC was enhanced in Pon3 KO kidneys (Fig. 7), which was likely a direct result of lower plasma [K⁺] in the Pon3 KO mice (Table 3) as opposed to a chaperone function of PON3. Both single cell RNAseq data and our previous IF staining study have detected higher levels of PON3 in ICs compared to PCs within the medulla (Shi et al., 2020). However, the abundance of the V-ATPase B1 subunit and pendrin, functional proteins that serve as markers of ICs, was not altered in kidney lysates from Pon3 KO mice as compared to littermate controls (Fig. 11), suggesting that neither protein is regulated by PON3’s chaperone function under baseline conditions. In addition, γENaC cellular distribution was altered specifically in CNT/CCDs of Pon3 KO kidneys (Fig. 5). Overall, our data suggest that PON3 many have specific targets within individual nephron segments or cell types within the kidneys, and the native targets of PON3 beyond ENaC require further exploration.

As discussed above, the regulatory effects of PONs on ion channels have been implicated in varied pathological conditions (Nagarajan et al., 2017; Pan et al., 2021; Hagmann et al., 2022). However, the renal phenotype of the Pon2 KO mice was only evident in an injury model (Hagmann et al., 2022). In agreement, we did not observe any abnormal kidney histology in Pon3 KO mice and the changes in Na⁺ and K⁺ handling were subtle in these mice at baseline. Future studies should explore the effect of PON3 deletion in renal injury models, such as Adriamycin-induced nephropathy, ischemia-reperfusion injury, or more chronic kidney disease models. One limitation of the current study is that the Pon3 deficient mice is a global KO model, and a loss PON3 expression may affect other organs, such as liver and lung. Nonetheless, our work provides the first insight of PON3’s physiological role in regulating Na⁺/K⁺ homeostasis in the distal nephron.

Biography

Stephanie Mutchler received her PhD from the University of Pittsburgh where she currently works as a postdoctoral fellow in the laboratory of Dr. Thomas Kleyman. Her work has explored the regulation of renal function by the hormone aldosterone. She hopes to continue addressing issues in renal care, both at a cellular level by increasing our understanding of mechanistic processes in the kidney as well as through the promotion of evidence-informed policy through her work as an incoming AAAS STPF fellow serving at the NIH NIDDK.