Renal FGF23 signaling depends on redox protein Memo1 and promotes orthovanadate-sensitive protein phosphotyrosyl phosphatase activity

Animal studies

Mice were fed standard chow (TS3242 Kliba Nafnag, Kaiseraugst, Switzerland) and were kept on 12/12 or 14/10 light–dark cycles. Mice floxed for exon 2 of the Memo1 gene (Haenzi et al. 2014) backcrossed to C57BL/6 J background over 10 generations were crossed with CreERTM transgenic mice (Hayashi and McMahon 2002) carrying a tamoxifen-inducible Cre recombinase controlled by a beta-actin promoter/enhancer. Genotypes were determined by PCR of ear punch biopsy DNA using primers: Memo forward 5′-CCCTCTCATCTGGCTTGGTA-3′, Memo reverse 5′-GCTGCATATGCTCACAAAGG-3′, Cre forward 5′-AGGTTCGTGCACTCATGGA-3′, Cre reverse 5′-TCACCAGTTTAGTTACCC-3′. For the current study, all animals used were males, with the exception of the experiments reported in Fig. 5 in which there was a sex ratio of 50:50 in each genotype. Loss of Memo was induced by 3 daily intraperitoneal injections with 2 mg tamoxifen at age 4 weeks (T5648 Sigma-Aldrich, distributed through Merck, Buchs Switzerland). Memofl/fl littermates without Cre but treated with tamoxifen served as controls. For FGF23 treatments, mice aged 6.5 weeks were starved for 6 h and intraperitoneally injected with 220 ng/g body weight of recombinant mouse FGF23 (2629-FG-025 R&D Systems, Minneapolis, MN, USA) in PBS-BSA 0.1% or PBS-BSA 0.1% vehicle with injection volume of 4 µL/g body weight. One hour later mice were dissected, a protocol adapted from (Gattineni et al. 2014). Treatments were randomly allocated by flipping a coin. Animals were euthanized after 1 h by terminal exsanguination under anesthesia. For the indicated pre-treatments to modulate FGF23 signaling, animals additionally received intraperitoneal injections of with sodium orthovanadate 20 mg/kg (13721-39-6 Sigma-Aldrich) (Wang et al. 2013) or sodium chloride 0.9% 1 h before their injection of FGF23 or PBS-BSA 0.1%.

Cell culture and transfection

HEK293 cells were obtained from ATCC and were tested negative for mycoplasma using LookOut® PCR-based kit (MP0035-1KT Sigma-Aldrich). HEK293-Klotho cells stably expressing Klotho were a gift from Dr. Bettina Lorenz-Depiereux, HelmholtzZentrum München and were described in Diener et al. (2015). Cells were transfected with Myc-DDK-tagged Rho-GDI1 (ARHGDIA) (MR202112 OriGene) using Lipofectamine 3000 (L3000015 Invitrogen, distributed through Thermo Fisher Scientific, Reinach, Switzerland). HEK293-Klotho cells were seeded at density of 500 000 cells/well on 6-well tissue culture plates and transfected with RhoGDI1 (ARHGDIA) (NM_004309) Human Tagged ORF Clone (RG200902 OriGene) using Lipofectamine 3000 Transfection Reagent (L3000015 Invitrogen, distributed through Thermo Fisher Scientific, Reinach, Switzerland) following the manufacturer’s protocol. 72 h after transfection, cells were treated with vehicle (0.1% BSA) or with 100 ng/ml FGF23 (SRP3039 Sigma-Aldrich, Buchs, Switzerland) for 5 or 15 min. Following treatment, cellular protein was extracted in RIPA (R0278 Sigma-Aldrich) by 1 h incubation on a shaker, followed by centrifugation at 10000 g.

Cellular siRNA experiments

Hek293-Klotho cells were seeded at density of 200 000 cells/well on 12-well tissue culture plates and transfected with ARHGDIA Ambion™ Silencer™ Select Pre-Designed siRNA (4427038 Ambion, distributed through Thermo Fisher Scientific, Reinach, Switzerland) or with RhoGDI1 (ARHGDIA) Human siRNA Oligo Duplex (Locus ID 396) (SR300287 OriGene Technologies Inc., Rockville MD, USA) using Lipofectamine 3000 Transfection Reagent (L3000015 Invitrogen, distributed through Thermo Fisher Scientific, Reinach, Switzerland) following the manufacturer’s protocol. 72 h after transfection with Rho-GDI1 or negative control siRNAs, cells were treated with vehicle (0.1% BSA) or with 100 ng/ml FGF23 (SRP3039 Sigma-Aldrich, Buchs, Switzerland) for 5 or 15 min. Following treatment, cellular protein was extracted in RIPA (R0278 Sigma-Aldrich) by 1 h incubation on a shaker, followed by centrifugation at 10000 g.

Redox proteomics

Kidney halves pooled from 2 mice per sample were homogenized in 10 mM Tris–HCl, pH 7.2, 250 mM sucrose, 1 mM EDTA, 150 mM KCl and 1 mM PMSF and spun down. Protein thiols and carbonyls were labelled with either 0.2 mM 5’-iodoacetamido fluorescein (IAF) or 1 mM fluorescein-5-thiosemicarbazide (FTSC), respectively, and incubated for 150 min at 37 °C in the dark. Proteins were precipitated with 20% TCA and centrifuged at 20,000 × g (3 min, 4 °C). Protein pellets were resuspended and washed with 100% ethanol/ethyl acetate (1:1) and 96% acetone, respectively, for carbonyl and thiol groups. Pellets were resuspended in Tris–HCl 0.5 M pH 6.8, glycerol 10%, SDS 0.5% and bromophenol blue and applied to SDS-PAGE. Gels were scanned for fluorescence in a Typhoon Trio Scanner 9400 (Control v5.0 + variable Mode Imager-RA 501: PRT < I/06/004, GE Healthcare, Buckingshamshire, UK; excitation, 490–495 nm; emission, 515–520 nm). Protein-associated fluorescence intensities (arbitrary units, AU) were analyzed using Quantity One image analysis software (BioRad, Hercules, CA, USA). Gels were stained with Colloidal Coomassie Brilliant Blue G250 and scanned.

For 2-dimensional analysis by gel electrophoresis (2DGE), proteins were first separated by pI (first dimension: isoelectric focusing IEF), followed by orthogonal separation according to molecular weight (second dimension: SDS-PAGE). Proteins were rehydrated in 5 M urea, 2 M thiourea, 2% CHAPS, 4% ampholyte (Pharmalyte 3–10, Amersham-Pharmacia Biotech, Little Chalfont, Bucks, UK), 1% Destreak reagent (Amersham-Pharmacia Biotech, Buckinghamshire, UK), and trace amounts of bromophenol blue, and then immobilized in 7 cm IPG strips (pH 3–10: 70 × 3 × 0.5 mm) and a linear gradient (NL) (GE Healthcare Immobiline™ Dry Strip IPG, GE17-6001–11 Bio-Sciences AB, Bio-Rad, Hercules, CA, USA). Proteins were focused at room temperature in a Protean IEF Cell (Bio-Rad) for at least 15 h, according to the following steps: (1) a linear voltage increase until 250 V for 15 min, (2) 10,000 V for 2 h (50 µA/ strip), (3) focusing at 20,000 V, and (4) hold at 500 V. Following IEF, strips were equilibrated for 20 min in equilibration buffer (6 M urea, 0.375 M Tris, pH 8.8, 2% SDS, 20% glycerol, containing 2% DTT) and then for 20 min in equilibration buffer containing 2.5% iodoacetamide. IPG strips were loaded onto 12% SDS-PAGE gels (PROTEAN Plus Dodeca Cell, BioRad). Gels were scanned for fluorescence as above and stained with Coomassie Blue followed by densitometry. Progenesis SameSpots Software (S/No.62605/3787; Nonlinear USA Inc, Durham, NC USA) was used to normalize FTSC/ IAF-labeled protein spots and Coomassie-staining intensity. Fluorescence spots were normalized to protein intensity for the same gel.

All experiments were performed in triplicates. Images of 2D gels were subjected to landmarking alignment so that corresponding spots were matched with each other, based on 3D Gaussian distribution after raw image correction and background subtraction. Spot intensities were normalized. Differences between protein spots in 2D gel images were automatically determined.

Spots with differences of p < 0.05 by ANOVA were manually excised from colloidal Coomassie-stained 2-DE gels. Proteins were extracted and enzymatically digested using a Perkin Elmer—Janus automated workstation. Following digestion, samples were reconstituted in 0.1% formic acid and analyzed using a Dionex U3000 Liquid Chromatography System (Dionex, Sunnyvale, CA, USA) and the Daltonics HCT Ion Trap Mass Spectrometer (Bruker, Glasgow, UK). The peptide fragment mass spectra were acquired in data-dependent AutoMS(2) mode with a scan range of 300–1500 m/z, and up to three precursor ions were selected from the MS scan (100–2200 m/z). Precursors were actively excluded within a 1.0-min window, and all singly charged ions were excluded. Following LC–MS/MS, Mascot Generic Files were created and the MASCOT database NCBInr was searched using Matrix Science webserver (www.matrixscience.com). The default search parameters were: enzyme = trypsin; maximum number of missed cleavages = 1; fixed modifications = carbamidomethyl (C); variable modifications = oxidation (M); peptide tolerance ± 1.5 Da; MS/MS tolerance ± 0.5 Da; peptide charge = 2 + and 3 + . Identified proteins were considered if a MASCOT score above 95% confidence was obtained (p < 0.05) and at least one peptide was identified with a score above 95% confidence (p < 0.05). This analysis was conducted at the Proteomics Core Facility of the University of Aberdeen, UK.​ All mass spectrometry proteomics data from these experiments have been deposited to the ProteomeXchange Consortium via the MassIVE repository with the dataset identifier PXD022342.

Cysteine oxidation analyses of recombinant protein

Amino acids 24-204 of human Rho-GDI1/ARHGDIA protein expressed in E. coli was obtained from Novusbio (NBP1-50861, Centennial, CO, USA NBP1-50861), with > 95% purity by SDS-PAGE and supplied in 20 mM Tris–HCl buffer (pH 8.0), 1 mM DTT, 10% glycerol without preservatives. Recombinant human Memo expressed in E. coli with > 95% purity, supplied in 20 mM Tris–HCl buffer, pH 8.0, 50% glycerol, 5 mM DTT, 300 mM NaCl, 2 mM EDTA was from Antibodies-online (ABIN2130536, Aachen, Germany). Copper-reducing oxidant activity of the batch of recombinant Memo protein used in this experiment was determined and reported in Moor et al. (2020).

First, as a putative metal cofactor pre-loading step, 2 µg Memo was incubated with RCM buffer with or without 10 µM CuCl2. After 30 min incubation at 4 °C the free CuCl2 was removed by Slide-A-Lyzer™ MINI Dialysis Device, 3.5 K MWCO (69550 Thermo Fisher Scientific) against RCM buffer for 30 min at 4 °C. Proteins were incubated for 15 min as 1000 ng Rho-GDI1 alone; 1000 ng Rho-GDI1 with 250 ng Memo; 1000 ng Rho-GDI1 with 250 ng CuCl2 pretreated Memo; and 1000 ng Rho-GDI1 with 100 µM of hydrogen peroxide as a positive control (H1009 Sigma-Aldrich), all in duplicates for each experiment, to individually label the total cysteine and the oxidized cysteine content of Rho-GDI1.

The samples for total cysteine content labeling were reduced with 5 mM Tris (2-carboxyethyl) phosphine (TCEP) (646547 Sigma-Aldrich). Total cysteine content was labeled with tandem mass tag (TMT) iodoTMTsixplex™ (90102 Thermo Fisher Scientific) followed by quenching with 20 mM 1DTT (43816 Sigma-Aldrich).

The samples for oxidized cysteine content labeling were treated with 500 mM iodoacetamide (I6125 Sigma-Aldrich) for 30 min followed by quenching with 20 mM DTT and reducing with 5 mM TCEP. The oxidized cysteine content of the samples was labeled with iodoTMTsixplex™.

Samples were precipitated with the SDS-PAGE Clean-Up Kit (10074304 GE Healthcare Life Sciences, distributed through Thermo Fisher Scientific) and separated on SDS-PAGE, followed by Coomassie Blue staining and manual band excision. Proteins were digested by chymotrypsin at 50 °C for 5 h and analyzed by LC–MS/MS (PROXEON coupled to a QExactive HF mass spectrometer, Thermo Fisher Scientific) with two injections of 5 μl digests. Peptides were trapped on a µPrecolumn C18 PepMap100 (5 μm, 100 Å, 300 μm × 5 mm, Thermo Fisher Scientific) and separated by backflush on a C18 column (5 μm, 100 Å, 75 μm × 15 cm, C18) by applying a 20-min gradient of 5% acetonitrile to 40% in water, 0.1% formic acid, at a flow rate of 350 nl/min. The Full Scan method was set with resolution at 60,000 with an automatic gain control (AGC) target of 1E06 and maximum ion injection time of 50 ms. The data-dependent method for precursor ion fragmentation was applied with the following settings: resolution 15,000, AGC of 1E05, maximum ion time of 110 ms, mass window 1.6 m/z, first mass 100 m/z, collision energy 27, under fill ratio 1%, charge exclusion of unassigned and 1 + ions, and peptide match preferred, respectively. Spectra were interpreted with Proteome Discoverer 2.4.0.305, with chymotrypsin rules allowing up to 8 missed cleavages, using variable modification of carboamidomethylated (+ 57.021 Da), dioxidation (+ 31.990 Da), trioxidation (+ 47.985 Da) and iodoTMT labelling (+ 324.216 Da) on Cys, and variable modification of oxidation (+ 15.99 Da) on Met, and acetylation (+ 42.011 Da) on protein N-Term. Parent and fragment mass tolerances were set to 10 ppm and 0.02 Da, respectively. Strict target false discovery rate for highly confident peptide-spectrum matches was set to 0.01. Protein identifications were only accepted when two unique peptides fulfilling the 1% FDR criterium were identified. This analysis was performed at the Proteomics Mass Spectrometry Core Facility of University of Bern. Resulting intensity peaks of specific peptide modifications were normalized by total signal intensity and are displayed as heatmap of non-transformed z-scores, i.e. a subtraction of the row mean followed by a division by the row standard deviation. All mass spectrometry proteomics data from recombinant protein analyses have been deposited to the ProteomeXchange Consortium via the PRIDE repository with the dataset identifier PXD022382.

Western blot

For immunoblotting, lysates of cells and tissues were lyzed in RIPA buffer or NP-40 buffer (150 mM NaCl, 50 mM HEPES pH7.4, 25 mM NaF, 5 mM EGTA, 1 mM EDTA, 1% Nonidet P-40, 2 M Na ortho-vandate and 1 mM DTT supplied with protease inhibitors leupeptin 10 μg/L, aprotinin 10 μg/L, and PMSF 1 mM) were prepared and denatured. Proteins were separated by SDS-PAGE and transferred onto nitrocellulose or PVDF, stained by Ponceau S, blocked in dried nonfat milk 5%-TBST or bovine serum albumin (A9647 Sigma-Aldrich) 3% in TBST before incubation with primary antibodies against Memo (1:2000, produced in-house (Haenzi et al. 2014) or 1:1000, HPA042603 Sigma-Aldrich), pERK (1:1000, sc-7383 Santa Cruz, Dallas TX, USA), tERK (1:1000, sc-93 Santa Cruz), actin (1:2000, A2066 Sigma-Aldrich), Rho-GDI1 (1:500, ABIN969501 Antibodies-online), Rac1 (1:500, ab33186 Abcam, Cambridge UK), RhoA (1:500, NBP2-22529 NovusBio) and Klotho (1:1000, AF1819 R&D Systems). Membranes were incubated with anti-mouse or anti-rabbit horseradish peroxidase-conjugated secondary antibodies (Milian Analytica, Rheinfelden Switzerland or ImmunoResearch, distributed through LubioScience GmbH, Zürich, Switzerland) and imaged using Fusion Solo (Witec AG, Sursee Switzerland) or a ChemiDoc XRS + System (BioRad). For reprobing, membranes were stripped using a low pH buffer (25 mM glycine–HCl, pH2, 0.4% (w/v) SDS).

Biochemical analyses

Serum electrolytes were analyzed by the Lausanne University Hospital: Total calcium (NM-BAPTA method), phosphate (phosphomolybdate method), and creatinine (modified Jaffé method). Intact FGF23 was analyzed using an ELISA (Kainos Japan CY-4000) following the manufacturer’s instructions. Protein phosphotyrosyl phosphatase (PTP) activity was measured in kidney lysates using fluorometric Protein Tyrosine Phosphatase Activity Assay Kit (#K829, BioVision Inc., Milpitas CA, USA) according to the manufacturer’s instructions. For quantification of Rho-GTPase activity in kidney homogenates, colorimetric Rac1 and RhoA G-LISA Activation Assay Kits (BK128-S and BK124-S Cytoskeleton Inc., Denver CO, USA) were used according to the manufacturer’s instructions.

RNA isolation

Frozen kidneys were homogenized with metal beads in a TissueLyser (Qiagen, Hombrechtikon, Switzerland). RNA was extracted using TRI Reagent Solution (AM9738 Ambion, Austin, TX, USA) for all downstream applications.

qPCR

One μg of RNA per kidney half was reverse-transcribed using PrimeScript RT Reagent kit (RR037 TAKARA, Shiga, Japan). Two μl of cDNA was used for quantitative real-time PCR to assess gene mRNA expression. Assays were performed using SYBR Green (Applied Biosystems, Foster City, CA UA) on a 7500 Fast machine (Applied Biosystems). Samples were run in triplicates in 20 uL total volume for each gene, and actin or GAPDH was used for normalization. Melting curves were obtained for every run. Program settings were: 95 °C 20 s, 40 cycles (95 °C 3 s, 60 °C 30 s), and for melting curve stage: 95 °C 15 s, 60 °C 1 min, rising at 1% ramp speed to 95 °C (15 s), 60 °C 15 s. Data were analyzed using the delta-delta CT method. Primers were ordered from Microsynth (Balgach, Switzerland), and sequences are shown in Supplemental Table 1.

RNAseq

For RNAseq, extracted RNA from mouse kidney halves was pairwise pooled to 3 individual pools for each of 4 experimental conditions (2 treatment states, 2 genotypes). Quality of mRNA was assessed and verified using a Fragment Analyzer (Advanced Analytical Technologies Inc.). Following quality validation, TruSeq Stranded RNA Sample Prep Kit (20020596 Illumina, San Diego CA, USA) was used with 1ug of total RNA for the construction of sequencing libraries. cDNA was sequenced using the Illumina HiSeq2500 platform and single-end reads (1 × 100). All sequencing data are accessible on National Center for Biotechnology Information, Sequence Reads Archive, Accession: PRJNA672305. Purity-filtered reads were trimmed with Cutadapt (v. 1.3, Martin 2011), filtered using seq_crumbs (v. 0.1.8) and aligned against the Mus musculus (Ensembl version GRCh38.82) transcriptome using STAR (v. 2.4.0f1). Data were normalized with the Trimmed Means of M-Values (TMM) method of the edgeR, and transformed with voom method of bioconductor package limma. Genes expressed at low levels were removed when none of the samples had more than 1 count per million reads. Only genes labeled as “protein coding” were retained for the statistical tests. A heatmap was constructed using the heatmap2 function in R 4.1.0 package gplots using z-scores of log-transformed normalized reads.

Statistical analyses

Analyses of parametric continuous variables were performed by t-test or by ANOVA, followed by post-tests using Bonferroni correction for multiple testing across all experimental groups. Analyses of nonparametric continuous variables were assessed by Kruskal Wallis test, followed by Dunn’s correction for multiple testing across all experimental groups. Analyses were performed in GraphPad Prism 5.0, and two-tailed p-values of < 0.05 were considered significant.

Analyses of RNAseq data were performed with limma and DESeq2 (R version 3.2.1, limma version 3.22.7 and DESeq2 version 1.8.1). Because the experiment has a 2 × 2 factorial design with two mouse genotypes and two treatment states, linear models were generated with all 4 groups as factors. Subsequently, contrasts of interest were extracted using moderated t-tests for the following groups: 1. "KOFGF-KOV" (treatment effect in Memo cKO), 2. "WTFGF-WTV" (treatment effect in floxed control genotype), 3. "KOFGF-WTFGF" (KO genotype effect in FGF23-treated mice), 4. "KOV-WTV" (KO genotype effect in vehicle-treated mice), and 5. "(KOFGF-KOV)-(WTFGF-WTV)" for the interaction between genotype and treatment. P-values were adjusted by the Benjamini–Hochberg method, controlling for false discovery rate (FDR) across all 5 contrasts as a correction of multiple testing conditions. All sequencing data are available on National Center for Biotechnology Information, Sequencing Reads Archive under the identifier PRJNA672305. Gene ontology enrichment was assessed using GOrilla, accessed on Nov 17th 2020 (Eden et al. 2009) using entire genes lists ranked by adjusted p-values.

Study approval

The present study was approved by the veterinary services of the Cantons of Vaud and Bern, Switzerland.

留言 (0)

沒有登入
gif