Mouse models
Experiments with mice were performed in accordance with institutional guidelines and with the approval of the Institutional Animal Care and Use Committees of Soochow University (Suzhou, China). The overall development and health of the animals were monitored according to the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) guidelines. Mice were housed in a specific pathogen-free (SPF) animal facility at Soochow University under a 12-hour light/dark cycle, a temperature range of 24 ± 2°C, and a humidity of 55 ± 5%. All mice were fed with acidified water and standard LabDiet® 5K52 diet formulated with 6.2% fat. Male mice of 12-week-old were used in all the experiments, unless otherwise specified. Littermates were used as controls in experiments unless specified. AlbCrePDIfl/fl mice and AlbCrePdia3fl/fl mice were generated by mating albumin-Cre (Albcre) mice with Pdifl/fl mice and Pdia3fl/fl mice41. The embryonic stem cells for PDI-floxed mice (clone number: EPD0317_6_D10), PDIA2 (PDIp) KO first (Pdia2−/−) mice (clone number: EPD0753_5_D11), PDIA5 (PDIr) KO first (Pdia5−/−) mice (clone number: DEPD00576_3_G10), PDIA8 (ERp27) KO first (Pdia8−/−) mice (clone number: EPD0688_4_C05), PDIA9 (ERp29) KO first (Pdia9−/−) mice (clone number: EPD0667_5_E05), PDIA13 (TMX3) KO first (Pdia13−/−) mice (clone number: HEPD0721_1_D05), PDIA14 (TMX4) KO first (Pdia14−/−) mice (clone number: EPD0684_1_C05), PDIA15 (ERp46) KO first (Pdi15−/−) mice (clone number: HEPD0564_9_A11), and PDIA19 (ERdj5) KO first (Pdia19−/−) mice (clone number: HEPD0721_1_D05) were generated by the International Knockout Mouse Consortium (IKMC) at the Cambridge-Suda Genomic Resource Center. After passing production quality control, the ES cells were injected into murine blastocysts and transferred to pseudopregnant female mice to generate the target KO first mice. PDIA16 (ERp18) whole-body knockout mice (CAGcrePdia16fl/fl) were generated by mating CAGcre mice with Pdia16fl/fl mice. Pdia16fl/fl mice were produced by Cyagen Biosciences Inc. by a CRISPR/Cas9-based protocol. Briefly, the gRNA to mouse Pdia16 gene, the donor vector containing loxP sites, and Cas9 mRNA were co-injected into fertilized mouse eggs to generate targeted conditional knockout offspring. ERp72(PDIA4) KO first (Pdia4−/−) mice 42, and TMX1(PDIA11) KO first (Pdia11−/−) mice 64 were generated and characterized as previously described. Genotyping of mice was performed by PCR analysis of tail DNA, and the gene expression was confirmed by semiquantitative reverse transcription-polymerase chain reaction (RT-PCR) or Western blotting. Primers used for genotyping and semiquantitative RT-PCR were listed in Table S3.
Cell lines
HepG2 (SCSP-510) and Hepa1-6 (SCSP-512) cells were obtained from the National Collection of Authenticated Cell Cultures (NCACC) in China. All cells were cultured in a Dulbecco’s modified Eagle medium (DMEM) containing 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin at 37℃ in the presence of 5% CO2.
Quantitative and semiquantitative RT-PCR
Mouse liver (20–50 mg) or hepatic cells (106 cells) were homogenized in 1 ml TRIzol reagent (Invitrogen, USA), and total RNA was extracted. First-strand cDNA synthesis was performed using the 1st Strand cDNA Synthesis kit (Vazyme) according to the manufacturer’s instructions. RT-qPCR was performed on a 7500 Real-time PCR system (Applied Biosystems) using a SYBR qPCR Master Mix (Vazyme). For semiquantitative RT-PCR, PCR products were analyzed in 1.5% agarose gels, stained with GelRed (Biotium), and photographed under ultraviolet light. The primers are listed in Table S3.
Western blotting and immunoprecipitation
Mice were euthanized and the liver was perfused with phosphate-buffered saline (PBS). The livers or hepatic cells were lysed with lysis buffer (20 mM Tris-HCl, 150 mM NaCl, pH 7.4, 1% Triton X-100, 1 mM EDTA, 0.5% sodium deoxycholate, and protease inhibitor cocktail form Roche) and centrifuged to discard the cell debris. Protein concentration was measured using a BCA Protein Assay Kit (Beyotime). For blotting of ApoB-100, the liver lysates or FPLC fragments were boiled with 2× Laemmli sample buffer containing 5% β-mercaptoethanol for 10 minutes at 70℃, for other proteins, the temperature was set at 100℃. Equal amounts of liver or cell lysates were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and transferred onto a polyvinylidene difluoride (PVDF) membrane. The membrane was blocked by 5% skimmed milk for 1 hour at room temperature and probed with various antibodies as indicated. Actin or GAPDH was used as loading control. For immunoprecipitation, the antibody was added into liver or cell lysate and incubated for 2 hours at 4℃, followed by incubation with protein G beads for 1 hours. The beads were washed with lysis buffer 4 times and boiled with 2× Laemmli sample buffer (Bio-Rad) containing 5% β-mercaptoethanol for 5 minutes at 100℃. The samples were analyzed by SDS-PAGE and Western blotting. The intensity of each band was quantitated using ImageJ software and normalized to the loading control.
Plasma biochemistry parameters and hepatic lipid analysis
Mouse blood was collected by venous puncture using EDTA anticoagulant and centrifuged at 2000g for 15 minutes to obtain plasma. Plasma levels of ALT, AST, TP, ALB, BILT, GLU, BUN, LDH, TG, CHOL, HDL, and LDL were measured using a fully automatic biochemical analyzer (Hitachi 7100). Hepatic levels of TG and CHOL were determined using the assay kits from EnzyChrom™ (Bioassay, Hayward, CA). Briefly, 400 mg of liver was homogenized in 4 ml of lipid extraction buffer (n-heptane/isopropanol = 4/7, v/v). After centrifugation, the supernatant was measured for the contents of TG and CHOL in livers using the assay kits according to the manufacturer’s instructions.
Histology and transmission electron microscopy
For histological analysis, liver tissues were excised and fixed in 4% paraformaldehyde. After paraffin embedding, the fixed specimens were sectioned at 5 µm and mounted on glass slides. The sections were then deparaffined, hydrated and stained with hematoxylin and eosin (H&E). For Oil Red O staining, the fixed livers were embedded in OCT compound followed by snap freezing. Frozen tissues were sectioned at 8 µm and stained with Oil Red O and hematoxylin according to the manufacturer’s instructions. For Toluidine blue staining, the fresh liver samples were quickly cut into blocks of 1 mm3 and fixed in 2.5% glutaraldehyde + 2.5% paraformaldehyde in 0.1MPBS pH 7.4, for 4 h at 4℃, rinsed, post-fixed with 1% osmium tetroxide at 4°C for 2 h. After wash with cold PBS, samples were dehydrated in graded concentrations of alcohol, and embedded in Epon812 (55℃, 48 h). Semithin sections (1.5 µm) were cut stained with 1% Toluidine Blue43, 44. Images were acquired using an orthostatic optical microscope (Nikon Eclipse E100) with an imaging system (NIKON DS-U3).
For transmission electron microscopy, mice were euthanized and perfused with PBS. A small piece of liver less than 1 mm3 was cut-off quickly and immersed in a fixative solution for TEM (Servicebio) at 4℃ for 4 hours. After dehydrating in different concentrations of ethanol solution, the liver was permeated in acetone for 30 minutes, in a 1:1 mixture of acetone and 812 embedding (Embed-812) for agents3 hours, in a 2:1 mixture of acetone and Embed-812 overnight, and in pure EMBed-812 for 6 hours. Embedding was performed at 60°C for 48 h, and ultrathin sections at 60 nm was cut using an ultramicrotome. Sections were stained with uranyl acetate in pure ethanol for 15 min, and leas citrate for 15 min, then dried overnight at room temperature. The images were collected and analyzed using a transmission electron microscope (HITACHI HT7700).
VLDL secretion assay and fast-protein liquid chromatography (FPLC) analysis
Mice fasted for 16 hours were anesthetized and injected i.v. with tyloxapol (150 mg/ml suspension in saline) at the dose of 500 mg/kg body weight. EDTA anticoagulated blood samples were collected at indicated timepoint (0, 2, and 4 hours after tyloxapol injection). Plasma was obtained by centrifugation of 2000g for 15 min at room temperature, and levels of plasma TG, CHOL, LDL and HDL were measured using a fully automatic biochemical analyzer as described above. For some experiments, the plasma samples were used for quantification of fatty acid esters by UHPLC-MS/MS, for other experiments, pooled plasma samples were used for fast-protein liquid chromatography (FPLC) analysis.
Quantification of fatty acid esters by UHPLC-MS/MS
The total contents of long chain fatty acids and fatty acid esters in liver and plasma were quantified based on UHPLC-MS/MS platform. For liver sample, 20 mg of flash-frozen liver spiked with 400µL water and deuterated fatty acid internal standard solution (Cayman Chemical) was homogenized using a tissuelyser with beads at 40 Hz for 4 min. The homogenates were mixed with 400 µL methanol and 800 µL chloroform, followed by vortex and centrifugation at 3500 g for 10 min. The chloroform phase was evaporated to dryness, redissolved in 500 µL of 75% ethanol and 0.5M KOH and incubated in water bath at 80°C for 1 h. After cooling, 100 µL formic acid and 600 µL hexane were added, vortexed, and centrifugated for 10 min. The hexane phase was diluted 40-fold and 20 µL was dried under nitrogen gas. The dried residues were reconstituted in 20 µL 1-hydroxybenzotriazole (HoBt), 40 µL cholamine and 20 µL 1-[Bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium 3-oxide hexafluorophosphate (HATU), and the mixture was incubated at room temperature for 30 min. After derivatization, 120 µL acetonitrile was added.
The UHPLC-MS/MS analysis was performed on an Agilent 1290 Infinity II UHPLC system coupled to a 6470A Triple Quadrupole mass spectrometry. Samples were injected onto a Hypersil™ GOLD C18 HPLC column (100mm × 2.1mm, 1.9µm, Thermo Scientific) at a flow rate of 0.3 mL/min. The mobile phase consisted of (A) water with 0.1% formic acid and (B) acetonitrile with 0.1% formic acid. The chromatographic separation was conducted by a gradient elution program as follows: 0 min, 10% B; 4 min, 30% B; 8min, 45% B; 11 min, 50% B; 14 min, 70% B; 15 min, 90% B; 15.5–18.0 min, 100% B; 18.1–20 min, 10%B. The eluted analysts were ionized in an electro spray ionization source in positive mode (ESI+). The temperatures of source drying gas and sheath gas were 300°C. The flow rates of source drying gas and sheath gas were 5 and 11 L/min, respectively. The pressure of nebulizer was 40 psi, and capillary voltage was 4000 V. The dynamic multiple reaction monitoring (dMRM) was used to acquire data in optimized MRM transition.
The raw data were processed by MassHunter Workstation Software (version B.08.00, Agilent). The peak areas of each compound in all samples were integrated. Calibration curves of ten-point were constructed by plotting the peak area ratio of each compound to internal standard against concentration of each compound. The concentrations (C, ng/mL) of fatty acids in prepared sample (for determination) were quantified automatically, and finally output for quantitative calculation of milligram liver samples of mouse in Excel with the following formula:
Content (ng/ mg sample) = 10×40×0.6×C/m
where C is the concentration quantified in prepared sample (ng/mL), m is the weight of tissue sample (mg).
For plasma sample, 20µL of mouse plasma was spiked with 220 µL of 0.1% formic acid in water, then the homogenates were mixed with 240 µL methanol and 480 µL chloroform, followed by vortex and centrifugation at 3500 g for 10 min, and the chloroform phase was evaporated to dryness. The following operation of derivatization and UHPLC-MS/MS analysis were undertaken as same as liver sample. The formula for quantitative calculation of milliliter plasma samples of mouse was as follows:
Content (µg/ mL plasma) = 10×30×C
where C is the quantified concentration in prepared sample (µg/mL).
TMT-LC-MS/MS-based quantitative proteome analysis of mouse liver samples
Quantitative proteomics was performed based on tandem mass tag-labeling combined with liquid chromatography-mass spectrometry/mass spectrometry (TMT-LC-MS/MS) technology. Protein extraction, in-solution trypsin digestion, HPLC fractionation, and LC-MS/MS analysis were performed as previously described65. Briefly, the livers isolated from 3 pairs of Alb-cre/Pdifl/fl and Pdifl/fl mice were mixed respectively and grinded in liquid nitrogen. After adding of four volumes of lysis buffer (8 M urea, 1% protease inhibitor cocktail, 3 µM treostatin A, 50 mM nicotinamide and 2 mM EDTA), the samples were sonicated three times on ice. After centrifugation at 12,000 g for 10 min at 4°C, the supernatant was collected and the protein concentration was measured. The samples were reduced with 5 mM dithiothreitol for 30 min at 56°C and alkylated with 11 mM iodoacetamide for 15 min at room temperature in darkness. After trypsin digestion, peptide was desalted, vacuum-dried and reconstituted in 0.5 M triethylammonium bicarbonate (TEAB). The Alb-cre/Pdifl/fl and Pdifl/fl samples were incubated with TMT6-126 and TMT6-127 for 2 hours at room temperature according to the manufacturer’s protocol, followed by mixture, desalination, and drying by vacuum centrifugation. The labeled peptides were fractionated into 18 fractions by high pH reverse-phase HPLC using an Agilent 300Extend C18 column and dried by vacuum centrifugation. Each fraction was dissolved in 0.1% formic acid and loaded onto a home-made reversed-phase analytical column (15-cm length, 75 µm i.d.) to separate the peptides, which were injected into a nanospray ionization (NSI) source for ionization and analyzed by a Q Exactive™ Plus Mass Spectrometer (Thermo Scientific). The electrospray voltage applied was 2.0 kV. The m/z scan range was 350 to 1800 for full scan, and intact peptides were detected in the Orbitrap at a resolution of 70,000. Peptides were then selected for MS/MS using normalized collision energy (NCE) that was set to 28 and the fragments were detected in the Orbitrap at a resolution of 17,500. Fixed first mass was set as 100 m/z. The MS/MS data were processed using the Maxquant search engine (v.1.5.2.8).
Isolation and culture of primary hepatocytes
Mice fasted for 12 hours were anesthetized and the peritoneal cavity was opened. The liver was perfused in situ through the portal veins at 37°C with 25 ml of Hank’s buffer containing 0.5 mM EGTA to remove the blood, and the perfusion was continued with 25 ml of digestion buffer (0.5 mg/ml collagenase IV and 15 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) in DMEM-low glucose medium) for 5 minutes. The liver was then excised and placed into a 10 cm dish containing isolation buffer (DMEM/F-12 containing 10% FBS). The lobes of the liver were teared apart, and the remaining section of the liver was shaken gently to free residual cells. After filtering through a 70-micron membrane, the cell suspension was washed four times with isolation buffer by centrifugation. The hepatocytes were suspended in isolation buffer and plated onto a type I collagen-coated 6-well plates (600000 cells each well). After incubation for 1 hour, the culture medium was replaced by DMEM low glucose medium with 10% FBS. Four hours later, the medium was replaced with a fresh DMEM-low glucose without serum. All the hepatocyte-based studies were completed within 48 hours post-plating.
MPB labeling and pull-down assay
The liver tissues isolated from Alb-cre/Pdifl/fl and Pdifl/fl mice were solubilized using lysis buffer (20 mM Tris-HCl, 150 mM NaCl, pH 7.4, 1% Triton X-100, 1 mM EDTA, 0.5% sodium deoxycholate, and protease inhibitor cocktail) and centrifuged to discard cell debris. Mouse hepatocytes lysates (2–3 mg/mL, 200 µL) were incubated with 25 µM MPB for 30 minutes at room temperature, followed by addition of 100 µM GSH to quench the labeling reaction. The lysates were incubated with streptavidin agarose beads with rotation overnight at 4℃. After washing with lysis buffer, the beads were boiled with 80 µL of 2× Laemmli sample buffer (Bio-Rad) containing 5% β-mercaptoethanol for 5 minutes. The samples were analyzed by SDS-PAGE and Western blotting.
Parallel reaction monitoring-mass spectrometry (PRM-MS)
To generate a spectral library, we synthesized 7 peptides covering 9 of the 10 cysteines in mature MTP protein, and a cysteine-free peptide that was most abundant in liver lysate detectable by DDA mode of mass spectrometry. Because the peptide with Cys298 acquired by trypsin digestion contained only 3 amino acids that was too short to be detected by MS, this peptide was not included. Briefly, 20 µg of synthetic standard peptide dissolved in 200 mM TEAB was reduced by 5 mM of DTT for 1 hour at 37℃, and alkylated with 15 mM iodoacetamide for 1 hour at room temperature in dark. After desalting and vacuum-dry, these peptides was reconstituted in 0.1% formic acid at the concentration of 0.5 µg/µl and analyzed using DDA mode of mass spectrometry. The raw files were searched using a quantitative proteomics software package (MaxQuant 1.6.15.0), and the processed data was imported into Skyline 20.1 software to construct a spectral library (Table S2).
Differential alkylation and PRM analysis of liver lysate samples were then performed. Briefly, livers from 3 pairs of Pdifl/fl and Alb-cre/Pdifl/fl mice were grinded in liquid nitrogen and lysed with four volumes of lysis buffer (1% SDS containing 1% protease inhibitor cocktail). After removal of the debris by centrifugation at 12,000 g for 10 min at 4°C, the supernatant was collected, and the protein concentration was determined. Equal amounts of total proteins from each sample were taken and divided into two equal aliquots. One aliquot (group 1) was incubated with 15 mM of IAM for 1 hour at room temperature in dark to alkylate the native reduced cysteines that did not form disulfide bond, followed by precipitation with four volumes of precooled acetone at -20℃ for 2 hours. After washing with precooled acetone three times, the protein precipitation was redissolved in lysis buffer and treated with for 1 hour at 37℃ to fully reduce the oxidized disulfide bond. The newly reduced cysteines were alkylated with 15 mM NEM for 1 hour at room temperature in dark. The other aliquot (group 2) was treated with 10 mM TCEP to reduce all the disulfide bond, and all of the cysteines were alkylated with IAM.
The alkylated proteins were precipitated and washed with precooled acetone. After the samples were redissolved in 200 mM TEAB and digested overnight using trypsin (1:50), the digested peptides were dissolved in solvent A (0.1% formic acid and 2% acetonitrile in pure water) and separated using an EASY-nLC 1000 UPLC system. The liquid gradient was comprised of an increase from 8–30% solvent B (0.1% formic acid in 90% acetonitrile) over 16 min, 30–40% in 6 min and up to 80% in 4 min then holding at 80% for the last 4 min, at a constant flow rate of 500 nl/min. The separated peptides were injected into a nanospray ionization (NSI) source for ionization and analyzed by a Q Exactive™ Plus mass spectrometer, in which the previously prepared acquisition method has been uploaded. The electrospray voltage was set to 2.1 kV. Precursor peptides and the secondary fragment ions were detected and analyzed by a high-resolution Orbitrap mass spectrometer. The range for full MS scan was set to 350–1405 m/z. The precursor peptides were detected in the Orbitrap at a resolution of 70,000. Peptides were then selected for MS/MS using NCE set to 27/35, and the fragments were detected in the Orbitrap at a resolution of 17,500. Automatic gain control (AGC) was set to 3E6 for full MS and 1E5 for MS/MS. The maximum injection time was set at 50 ms for full MS and 200 ms for MS/MS. The isolation window for MS/MS was set at 1.6 m/z.
The MS data were processed using the Skyline software. The identities of the peptides (spectral matching) were confirmed by post-acquisition data processing. The traces of the qualified fragment ions of the peptides were extracted and the peak areas were determined and used for quantifying peptide abundance. The percentage of reduced form of each cysteine was calculated by comparing the peak area of matching peptide in group 1 to that in group 2. The peak area of the cysteine-free peptide was used as an internal standard to normalize the quantification of target peptides.
Expression of the recombinant MTP protein in HepG2 cells
The plasmid expressing full-length human MTP cDNA in a pCMV3 vector was purchased from Sino Biological Inc. FLAG-tag sequence DYKDDDDK was subcloned into the plasmid using an Ultra One Step Cloning Kit (Vazyme) for expressing the MTP protein at the C-terminus. The MTP mutants including C174A-C194A, C298A-C301A, C440A-C445A, C827A-C878A, C397A, and C866A were constructed using a Fast Mutagenesis Kit V2 (Vazyme). The primers used for homologous recombination and site-directed mutagenesis are shown in Table S3. Wild-type or mutant MTP plasmids were transfected into HepG2 cells using the Lipofectamine 3000 according to the manufacturer’s instructions. The cells were harvested for 48 hours after transfection, and the expressions of wild-type MTP and the variants were detected by Western blotting with anti-FLAG antibody.
FoldX analysis of MTP variants
The changes in Gibbs free energy (∆∆G) induced by the MTP mutations at each disulfide bonds were calculated using FoldX 66. Briefly, the Gibbs free energy (∆G) of wild-type MTP and mutant MTP was calculated based on the sum of the total intermolecular and intramolecular forces, determined by the FoldX forcefield. The relative ∆∆G was obtained by subtracting ∆Gmutant from ∆Gwild−type of MTP. The “buildmodel” function was used to build the 3D structure of the mutants based on the structure of wild-type MTP complex (PDB identifier 6I7S), and the structures of both wild type and the mutants were added into GeoPPI to obtain the corresponding predicted values67.
Molecular dynamics (MD) simulations
Simulations were prepared from the published MTP crystal structure (PDB identifier 6I7S). All MD simulations of each redox state (Cys174-Cys194, C298-C301, Cys440-Cys445, and Cys827-Cys878 and none of which was reduced) were performed with the GROMACS 2018 software package, using the parameter sets of the GROMOS 54A7 force field. All the simulations were performed under a SPC water model and periodic boundary conditions, allowing for a minimum distance of 1 nm from the protein to any box edge. The workflow of MD simulations includes four steps: energy minimization, heating, equilibration, and production. First, energy minimization was performed using 5000 steps of steepest descent algorithm, and 5000 steps of conjugate gradient algorithm while keeping the protein backbone atoms completely constrained. Next, the minimized structures were thermalized to 300 K over 50 ps. Then, the thermalized systems were simulated under isothermal-isobaric NPT ensemble over 50 ps to ensure equilibration of the periodic box volume. Finally, production dynamics were performed for 100ns in the NPT ensemble with a 2-fs time step and coordinates were saved every 40 ps for analysis.
Plasma characterization and FPLC analysis
After 16 h of fasting, 12-week-old Pdifl/fl and Alb-cre/Pdifl/fl male mice were injected i.v. with tyloxapol at a dose of 500 mg/kg body weight. Two hours later, blood was collected with EDTA as an anticoagulant, and plasma was isolated. A total volume of 300 µL pooled plasma samples of the same genotype were subjected to FPLC by using the Superose 6 Increase 10/300 GL columns at 0.5 mL/min flow rate. The plasma samples were fractionated into 40 tubes, each containing 300 µL of volume68, 69. Total triglycerides and cholesterol levels of each fraction were measured by commercial kits (100000220 and 100000180 from Zhongsheng Beikong, respectively) according to the manufacturer’s protocol. The level of ApoB protein in each fraction was detected by Western blotting.
PDI reductase activity assay and MTP activity assay
Fifty milligrams of liver tissue were homogenized in 1 mL of 10 mM Tris-HCl, pH 7.4 buffer containing 150 mM NaCl, 1 mM EDTA, 0.5 mM PMSF and 20 µg/ml leupeptin and then sonicated on ice. After centrifugation at 10000 g for 10 min at 4℃, the intermediate layer of the supernatants was collected. Repeat the above steps twice. After protein quantification, 100 µg of liver lysates were used for PDI activity assay and MTP activity assay according to the instructions provided by the manufacturers (ENZ-51024 and Roar#RB-MTP, respectively) 25, 34, 38, 70.
Quantification and statistical analysis
No statistical method was used to predetermine the sample size. Specifically, at least triplicates were included for any statistical analysis. Data analysis was performed using the GraphPad Prism 8 software. For parametric comparison, one-way ANOVA for multiple groups and Student’s t-test for two groups were used, and the values were expressed as the mean ± SEM. P < 0.05 was considered statistically significant. Asterisks denote corresponding statistical significance *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 (two-tailed Student’s t-test).