Assessment of lipotoxic endoplasmic reticulum (ER) stress in nonalcoholic steatohepatitis (NASH)

Abstract

Hepatocyte lipotoxicity is a hallmark of nonalcoholic steatohepatitis (NASH), and lipid induced liver injury occurs, in part, via activation of endoplasmic reticulum (ER) stress. Consequently, the unfolded protein response (UPR) is initiated, driven by three key ER transmembrane proteins, resulting in downstream responses that are dynamic and interconnected. Thus, careful interrogation of these pathways is required to investigate the complex role of ER stress in NASH. Herein, we describe different mechanisms of and in vitro assays for assessment of lipotoxic ER stress in mouse hepatocytes.

1. Introduction

The unfolded protein response (UPR) is a conserved endoplasmic reticulum (ER) to nucleus signaling response triggered by diverse conditions that interfere with ER homeostasis [1]. Two broad categories of conditions can activate the UPR, termed proteotoxic ER stress and lipotoxic ER stress [2]. Proteotoxic ER stress occurs when there is an interference with the protein folding capacity of the ER due to an excess of unfolded proteins or disruption of protein folding pathways [3]. Lipotoxic ER stress is best described in the context of the accumulation of saturated fatty acids such as palmitic acid (PA, C16:0) and myristic acid (C14:0), and sphingolipids [4] [5] [6,7]. The pathophysiological process of lipotoxic ER stress is highly relevant to cellular injury in obesity-associated disorders such as nonalcoholic steatohepatitis (NASH) in which hepatocytes are damaged due to the accumulation of toxic lipids including PA. Here we describe the key methodology to measure activation of lipotoxic ER stress in PA-treated hepatocytes (Table 1).

Table 1.

Methods to assess lipotoxic ER stress in primary mouse hepatocytes

1. UPR arms	Method	Notes	Ref
1.IRE1α
1.1.IRE1α Phosphorylation	Western blot	Concentrate the protein by immunoprecipitation prior to immunoblotting Use antibodies specific for the phosphorylated form	[24]
1.2.Xbp1 splicing	RT-PCR and qPCR	Use of PCR primers specific for the spliced and total Xbp1 mRNA	[25]
1.3.XBP1 target genes – ERdj4, Edem1, p58IPK,	qPCR
2.PERK
2.1.PERK phosphorylation	Western blot	Concentrate the protein by immunoprecipitation prior to immunoblotting Use of antibodies for total and phospho-specific proteins Use of appropriate positive controls (for e.g., Tunicamycin) as well negative controls (PERK inhibitors)	[9]
2.2.eIF2α phosphorylation			[26]
2.3.ATF4	Western blot		[27]
2.4.ATF4 target genes – CHOP, GADD34,	Western blot qPCR		[24], [28]
3.ATF6
3.1.ATF6 nuclear translocation	Reporter plasmid with 5 ATF6 sites that is highly sensitive to ER stress	Ensure appropriate concentration as overexpression of ATF6 can promote activation	[29]
3.2.ATF6 target genes – eg., Grp94, ERp57, ERp72, BiP	qPCR		[30,31]
2. ER morphology	Electron microscopy	Ultrastructural anatomy may be distorted during fixation of lipid laden hepatocytes	[9]
3. ER function – protein aggregation	Fluorescence	Thioflavin T	[32]

Three ER transmembrane proteins, inositol requiring enzyme 1α (IRE1α), protein kinase R-like ER kinase (PERK) and activating transcription factor 6α (ATF6α), are the canonical sensors of the UPR. The ER luminal domain of these proteins’ senses unfolded proteins; whereas lipotoxic ER stress activates the UPR sensors via changes in membrane saturation via a mechanism dependent on the transmembrane domain of the UPR sensors[8] [9]. Regardless, once activated, the UPR sensors unleash conserved signaling pathways which can be measured as a surrogate readout of ER stress. Recent studies have demonstrated that the transcriptional response activated in response to lipotoxic ER stress is distinct from proteotoxic ER stress [8,10], suggesting either selective context-dependent activation of a UPR-triggered transcriptional response, or the activation of additional pathways under lipotoxic conditions, which regulate the aggregate transcriptional response of the UPR.

Lipotoxic ER stress is best described in the context of several interrelated lipid species. These include saturated free fatty acids, sphingolipids, and lysophosphatidyl choline (LPC). PA is physiologically abundant, and also significantly elevated in obese subjects with nonalcoholic steatohepatitis (NASH) [11]. It has been well studied in activation of the UPR in concentrations typically ranging from 200–400 μM [4]. Albumin is included in cell culture conditions as a vehicle to bind PA and maintain the concentration of unbound PA in cell culture medium. We typically employ 1% albumin w/v (151 μM) and 400 μM PA to not exceed the solubility of PA [12][13]. PA is a substrate for de novo ceramide synthesis, which occurs in the ER [4]. Several ceramide-derivates can activate the UPR, and myriocin, an inhibitor of de novo ceramide synthesis, can mitigate PA-induced UPR, suggesting that ceramides partly mediate PA-induced activation of the UPR [6]. Similarly, intracellular LPC concentration increases in PA treated hepatocytes with LPC-induced activation of the UPR signaling pathways [14].

2. Materials

2.1. Isolation and treatment of primary mouse hepatocytes

1. Wild-type mice of C57BL/6 strain background, male, 3 months old.

2. Calcium-free Hanks’ Balanced Salt solution (HBSS)

3. Collagenase D perfusion solution: Add 35 mg of collagenase D to 125 mL of HBSS with calcium and magnesium.

4. Krebs-Henseleit buffer: Add Krebs-Henseleit Buffer powder containing 2 g D-glucose, 0.141 g magnesium sulfate, 0.16 g potassium phosphate monobasic, 0.35 g potassium chloride, 6.9 g sodium chloride to 1 L water.

5. Percoll solution: Add 22.5 mL Percoll stock solution to 2.5 mL of 10X phosphate buffered saline (PBS).

6. 0.4% Tryptan blue.

7. Supplemented Waymouth’s medium: Add 10% fetal bovine serum (FBS), 100,000 IU/L penicillin, 100 mg/L streptomycin, and 100 nM insulin to 1L of Waymouth’s medium.

8. Primary mouse hepatocyte (PMH) culture medium: Add 25 mM glucose, 100,000 units/L penicillin, 100 mg/L streptomycin, and 10% FBS to 1L of Dulbecco’s modified Eagle’s medium (DMEM).

9. Collagen coating solution, 50 mL.

10. Palmitic acid stock solution: Dissolve 100 mg of palmitic acid in 4.87 mL of isopropyl alcohol to obtain a stock concentration of 80 mM.

11. Palmitic acid treatment solution: Add 5 mL of PA stock solution and 10 g of fatty acid free bovine serum albumin (BSA) to 1L of DMEM to assure a physiological ratio between bound and unbound PA in the solution [16].

2.2. RNA and protein analyses

1. TRIzol^™ Reagent (ThermoFisher Scientific).

2. RNeasy Plus Micro Kit (QIAGEN): contains gDNA-eliminator columns, RNeasy spin columns, 2 mL collection tubes, buffer RW1 and buffer RPE.

3. NanoDrop ND1000 spectrophotometer (ThermoFisher Scientific).

4. iScript cDNA synthesis kit (Bio-Rad Laboratories).

5. QIAxcel instrument for capillary electrophoresis (QIAGEN).

6. Radioimmune precipitation assay (RIPA) buffer: Add 5 mL of 1M Tris HCl (pH 7.4), 1 mL of NP-40, 0.5 mL of 20% Sodium Deoxycholate (SDS), 3 mL of 5M NaCl, 200 μL of 0.5M EDTA to 50mLwater.

7. cOmplete^™, Mini Protease Inhibitor Cocktail.

8. Bio-Rad detergent compatible (DC) protein assay (Bio-Rad).

9. Transfer buffer: Add 14.4 g of Tris base and 3.03 g of Glycine to 500 mL of water, then add 200 mL of methanol to make 1 L of transfer buffer composed of 25 mM Tris, 192 mM, and 20% methanol.

10. PVDF Membrane for western blot.

11. Blocking buffer: 5% non-fat dairy milk (NFDM) in PBS or 5% BSA in PBS.

12. Appropriate primary antibodies against IRE1α, XBP1, p-IRE1 (pS724), PERK, p-PERK (Thr982), eIF2α, p-eIF2α, ATF4, CHOP, GADD34 and ATF6.

13. Horseradish peroxidase (HRP)-coupled secondary antibodies.

14. Chemiluminescent detection system.

15. β-Glycerophosphate disodium salt hydrate.

3. Methods

3.1. Isolation of primary mouse hepatocytes

1. Isolate mouse hepatocytes from C57BL/6 mice by a 2-step collagenase perfusion technique.

2. First, perfuse the liver in situ with approximately 40 mL of calcium-free HBSS at 4 mL/minute

3. Subsequently, perfuse with 125 mL of collagenase D perfusion solution.

4. Excise and mince the livers in Krebs-Henseleit buffer [15].

5. Centrifuge at 50 × g for 2 minutes at 4°C.

6. Aspirate supernatant, gently resuspend cells by inverting the tube in 30 mL cold DMEM and repeat the above centrifugation.

7. Discard the supernatant and resuspend in 25 mL of cold DMEM with 5% FBS.

8. To the dispersed cell suspension, add 25 mL of 1X Percoll solution.

9. Gently mix and centrifuge at 150 × g for 10 minutes at 4°C, with no brake.

10. Aspirate everything above the pellet and discard. The pellet contains hepatocytes.

11. Ensure viability of cells to be 95% by Trypan blue exclusion.

12. Plate primary mouse hepatocytes initially in supplemented Waymouth’s medium for 4 hours [14].

13. Subsequently, culture hepatocytes in in PMH culture medium [15].

14. Use collagen coated plates, and plate in 10 cm dishes, at an approximate density of 3.2×10⁴ cells/cm².

15. Attach cells for 6–12 hours prior to treatment with PA.

3.2. Treatment with palmitic acid

Incubate cultured cells with PA treatment solution. The concentration range of PA used in the experiments is 200–400 μM to match fasting PA concentrations observed in plasma in human NASH. The final concentration of the vehicle, isopropyl alcohol, is 1%. The duration of treatment with palmitic acid can be adjusted based on the readout of ER stress being studied (Figures 1 and 2A) (See Notes 1 and 2).

Figure 1. Suggested timeline for assessing lipotoxic ER stress in mouse hepatocytes in vitro.

After induction of lipotoxicity in vitro by application of palmitic acid, different readouts of ER stress can be examined at different time points as indicated. Readouts depicted in this figure can be assayed with techniques listed in Table 1.

Figure 2. Representative examples of ER stress responses to palmitic acid in primary mouse hepatocytes.

(A) Western blotting for expression of BiP, ATF4, phospho-eIF2α and CHOP in an immortalized mouse hepatocyte (IMH) cell line treated with vehicle (isopropyl alcohol), palmitic acid (400 μM), and tunicamycin (1.2 μM) as a positive control, for the indicated time periods.

(B) QIAxcel capillary electrophoresis images of total and spliced Xbp1 in primary mouse hepatocytes treated with vehicle (isopropyl alcohol), thapsigargin, tunicamycin and palmitic acid; 18s is included as the housekeeping control.

3.3. Extraction of RNA and quantitative real-time PCR

1. After treatment with PA, wash the cells twice with PBS and resuspend in 1 mL Trizol.

2. Homogenize the sample by pipetting up and down several times and leave at room temperature for five minutes.

3. Add 200 μL of chloroform, mix well, and centrifuge at 12,000 × g for 15 minutes at 4°C.

4. Carefully pipette up to 350 μL of the clear aqueous phase on the top and transfer to a gDNA-eliminator column placed in a 2 mL collection tube.

5. Centrifuge at 8000 × g for 30 seconds at 4°C. Discard the column and save the flow through.

6. Add equal volume (350 μL, as above) of 70% ethanol to the flow-through and mix well by pipetting. Do not centrifuge and proceed immediately to the next step.

7. Transfer the entire 700 μL of sample to a RNeasy spin column placed in a 2 mL collection tube.

8. Centrifuge at 8000 × g for 15 seconds at 4°C and discard the flow-through.

9. Add 700 μL Buffer RW1 to the RNeasy spin column, and centrifuge at ≥8000 × g for 15 seconds at 4°C. Discard the flow-through.

10. Add 500 μL Buffer RPE to the RNeasy spin column, and centrifuge at ≥8000 × g for 15 seconds at 4°C. Discard the flow-through.

11. Add 500 μL Buffer RW1 to the RNeasy spin column, and centrifuge at ≥8000 × g for 2 minutes at 4°C. Discard the flow-through.

12. Place the RNeasy spin column in a new 1.5 ml collection tube. Add 30–50 μL RNAse-free water to the spin column membrane and centrifuge at ≥8000 × g for 1 minute at 4°C to elute the RNA.

13. Assess quantity and quality of RNA spectrophotometrically.

14. Reverse transcribe 1000 ng of RNA per sample into cDNA by the iScript cDNA synthesis kit.

15. Quantitative real-time PCR reactions can be run with cDNA diluted at 1:20 to 1:40 in molecular grade water, with appropriate primers (Table 2).

Table 2.

List of primers for PCR

Gene	Forward primer	Reverse primer
18s	CGCTTCCTTACCTGGTTGAT	GAGCGACCAAAGGAACCATA
Xbp1 PCR	ACACGCTTGGGAATGGACAC	CCATGGGGAGATGTTCTGGG
Total Xbp1	AAGAACACGCTTGGGAATGG	ACTCCCCTTGGCCTCCAC
Spliced Xbp1	GAGTCCGCAGCAGGTG	CCATGGGGAGATGTTCTGGG
ERdj4	GGGAAGGATGAGGAAATCGTC	CACGAAACGCTTCCCCAT
Edem1	TTCCCTCCTGGTGGAATTTG	AGGCCACTCTGCTTTCCAAC
p58IPK	ACGCCTTTGACGGTGCCGATTA	AAGTCGCTGATGGCTTTCCTGG
Chop	CTGCCTTTCACCTTGGAGAC	CGTTTCCTGGGGATGAGATA
GADD34	CCCGAGATTCCTCTAAAAGC	CCAGACAGCAAGGAAATGG
Grp94	AATAGAAAGAATGCTTCGCC	TCTTCAGGCTCTTCTTCTGG
ERp57	CGCCTCCGATGTGTTGGAA	CAGTGCAATCCACCTTTGCTAA
ERp72	TCCCATTGCTGTAGCGAAGAT	GGGGTAGCCACTCACATCAAAT
BiP	TGTTCAACCAATTATCAGCAAACTC	TTCTGCTGTATCCTCTTCACCAGT

3.4. Capillary electrophoresis

1. After RT-PCR amplification of spliced and total XBP-1, resolve the amplicons by capillary electrophoresis on a QIAxcel instrument (Figure 2B) (see Notes 3 and 4).

2. Remove the QIAxcel DNA High Resolution Cartridge from 4°C and carefully place it in the QX cartridge Stand. Allow the cartridge to stand for a minimum of 20 minutes.

3. Rinse the buffer tray with hot water and fill the WP and WI positions of the buffer tray with 10 mL QX Wash Buffer, and cover with 2 mL mineral oil to prevent evaporation.

4. Bring the QX DNA Size marker (50bp-800bp) to room temperature and dilute it 1:40 using the DNA dilution buffer in a PCR tube.

5. Add 20 μL of mineral oil to the top of the PCR tube, and load this in position A12.

6. Bring the QX Alignment Marker (15 bp-3kbp) to room temperature, and load 15 μL into each tube of a QX 0.2 mL 12-Tube Strip.

7. Add 20 μL of mineral oil to each tube and place the strip into the MARKER1 position of the buffer tray.

8. Remove the QIAxcel DNA gel cartridge from the QX Cartridge Stand, and place into the instrument.

9. Insert the smart key into the smart key socket in any orientation and close the cartridge door.

10. Load a 96-well plate containing samples onto the sample tray holder. Ensure each well in each row you run is filled. If there are blanks, use dilution buffer with 20 μl of mineral oil laid on top.

11. On the QIAxcel ScreenGel software, select the “DNA high res” process profile from the drop-down list, and click next.

12. On the next screens, select samples on input sample information. Confirm the run details and start the run.

13. After the run, a window containing the gel image is displayed. Analysis of the DNA fragments in reference to the QX DNA Size Marker and an electropherogram can be visualized on the software.

3.5. Extraction of protein and Western blot analysis

1. After treatment with PA, collect the cells by scraping and lyse in RIPA buffer with protease as well as phosphatase inhibitors (see Note 5).

2. Measure protein concentrations by the Bio-Rad detergent compatible (DC) protein assay.

3. Load equal amounts of protein onto 4–15% gradient reducing sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS/PAGE).

4. After electrophoretic separation, transfer to a polyvinylidene difluoride (PVDF) membrane in a transfer buffer.

5. Block the membrane in PBS with 5% non-fat dairy milk (NFDM) for nonphosphorylated proteins and for PBS with 5% BSA for phosphorylated proteins.

6. Incubate overnight at 4° C with the appropriate primary antibodies against IRE1α, XBP1, p-IRE1 (pS724), PERK, p-PERK (Thr982), eIF2a, p-eIF2a, ATF4, CHOP, GADD34 and ATF6 (see Notes 6 and 7).

7. Primary antibodies should be diluted in PBS with 5% NFDM or with 5% BSA for phospho-specific antibodies.

8. Wash the membranes and then incubate with corresponding secondary horseradish peroxidase (HRP)-coupled antibodies and develop with a chemiluminescent detection system.

9. For quantification, densitometry can be performed using Image J (National Institutes of Health).

10. For each experiment the protein of interest can be normalized to its loading control and then expressed relative to vehicle control-treated conditions. See Notes 8–10.

4. Notes

There are several important considerations in the biochemical and molecular analysis of UPR signaling. Some of these are discussed below:

1. Kinetic analysis of the activation of ER stress response and UPR signaling should be performed. Some of the more proximal signaling pathways are activated early, some are sustained over time, and others are activated in a delayed manner. For PA-induced ER stress the kinetic profile for assessment of ER stress is shown in Figure 1.

2. Positive controls for activation of ER stress should be included. These include pharmacological agents like tunicamycin, thapsigargin, dithiothreitol (DTT), or brefeldin A. We typically include tunicamycin at a concentration of 0.6 – 3 μM and thapsigargin at a concentration of 0.1 μM, and perform assays for ER stress over time, e.g., at 1, 2, 4 and 8 hours (Figure 2A).

3. Due to the 26 base pair difference in size of the two PCR products, polyacrylamide gel electrophoresis in Tris-borate EDTA (TBE) buffer followed by detection of DNA with a nucleic acid dye (SYBR^™ Green) may yield better resolution than agarose gel electrophoresis if capillary electrophoresis is unavailable.

4. Ten μg of nuclear extract can be used to detect spliced XBP1 by western blotting, usually a single band is obtained at approximately 55 kDa.

5. The use of serine-threonine phosphatase inhibitors in the lysis buffer such as beta-glycerophosphate is recommended to enable measurement of phosphorylated proteins.

6. CHOP is expressed at a very low level basally, is induced under conditions of stress, and can be induced by other stress kinases, such as c-Jun N-terminal kinase [17]. Furthermore, CHOP antibodies are notoriously non-specific, and we use the commercially available antibody from Santa Cruz Biotechnology [18]. Therefore, for assessment of CHOP induction we recommend qRT-PCR for the mRNA and Western blotting for protein expression with the inclusion of positive and negative controls. CHOP knockout mice are commercially available (Jackson Laboratory) and should be used to generate a negative control tissue or primary mouse hepatocyte lysate. We do not typically isolate nuclear extract for assessment of CHOP induction and can detect it in whole cell lysate.

7. ATF6α antibodies have low selectivity for the endogenous protein and may yield non-specific bands [19]. The inclusion of negative and positive controls is recommended. AT6α floxed mice are commercially available (Jackson Laboratory) and should be used to generate positive and negative control tissue lysates.

8. Reporter constructs and dominant negative constructs are available for many of the key molecules that mediate the UPR [20–22]. Cell lines expressing these reporter constructs can be employed for screening, for example, to identify lipids that can induce ER stress, and mechanistic experiments.

9. Pharmacologic inhibitors of the UPR mediators can be employed to interrogate specific pathways. These include the IRE1α inhibitors (4μ8C, STF-083010), ATF6μα inhibitors (ceapins, nelfinavir), and eIF2α dephosphorylation inhibitor (salubrinal).

10. Protein phosphorylation status can be alternatively assessed with the Phos-Tag^™ acrylamide gel electrophoresis approach which separates out phosphorylated from nonphosphorylated proteins based on electrophoretic mobility [23]. Following transfer to PVDF membranes the proteins of interest can be detected by western blotting as described above.