Experimental procedures in animals
C57BL/6 J male mice were purchased from Jackson Laboratory (Bar Harbor, ME) or bred at the AAALAC-accredited University of Vermont or CIC bioGUNE from breeding pairs obtained from Jackson Laboratory. MCJ-deficient mice (MCJ KO) were already generated5. Leptin receptor mutant (Leprdb/J) mice were obtained from Jackson Laboratories. The studies were performed with male mice and experiments were initiated when the mice were between 10 and 15 weeks of age. Animal procedures were approved by the University of Vermont and University of Colorado Animal Care and Use Committee according to the criteria outlined in the Guide for the Care and Use of Laboratory Animals and institutional IACUC review committee. Animal work at CIC bioGUNE was approved by the institutional IACUC and the Competent Authority following Spanish and European regulations.
For fasting studies, all food was removed and tissues were harvested 36 h later.
For studies using the methionine and choline-deficient (MCD) diet model, mice were fed a diet completely deficient in methionine and choline (MCD) (Research Diets A02082002BRi) for either 2 or 3 weeks as indicated.
For studies using the modified methionine-,choline-deficient (mMCD) diet model, mice were fed a choline-deficient diet containing 0.1% methionine with 45 kcal% fat (Research Diets, A06071309i) for 4 or 6 weeks.
For studies using the HF-HFD model, mice were fed a high fat/high fructose diet consisting of a 58% fat diet with sucrose (Research Diets D12331i) in combination with water containing 42 g/L of carbohydrate (55% fructose and 45% sucrose)23. After 4 months on the diet, mice were treated weekly with siMCJ formulated with Invivofectamine i.v. or PBS for 5 weeks.
For studies using the CD-HFD model, mice were fed a choline-deficient, high-fat diet (45 kcal% fat, Research Diets, D05010402i) for 4 months prior to the initiation of GalNAc-formulated siMCJ treatment. Mice then received s.c. injections of 10 mg/Kg GalNAc siMCJ381or PBS every 2 weeks for a total of 4 doses.
For the studies using the diet induced obesity (DIO) mouse model and insulin tolerance test (ITT), B6J DIO male mice were purchased from Jackson Laboratory after 13 weeks in the D12492 high fat diet containing 60 kcal% fat. Mice were maintained with the same diet during the treatment. Mice were treated weekly with siMCJ formulated with Invivofectamine i.v. or PBS for only 2 weeks. Insulin tolerance test was performed a week after the last dose of siMCJ. For ITT, blood glucose was determined using a glucometer prior to the injection of insulin (time 0), following by the administration of insulin (0.75 mU/g) and determination of blood glucose after 15, 30, 60, and 90 min using the OneTouch Ultra glucometer and glucose strips.
For the studies using the carbon tetrachloride (CCl4) model of liver fibrosis, mice were administered i.p. 0.6 μl/g of CCl4 every 3 or 4 days. A group of mice were harvested 2 weeks after the first CCl4 dose. After 2 weeks on CCl4, mice were treated with invivofectamine/siMCJ (1.7 mg/Kg) or PBS control group weekly for 3 more weeks. Mice were then euthanized and tissues were harvested.
For treatment with siRNA in vivo, MCJ-specific siRNA (siMCJ) (position 294–312) was designed as previously described18. For in vivo delivery using Invivofectamine, siMCJ was complexed with lipid nanoparticle Invivofectamine® 3.0 (ThermoFisher Scientific) according to the manufacturer’s instruction. Mice received weekly i.v. injections of 1.7 mg/Kg of siRNA complexed with invivofectamine. MCJ-specific siRNAs formulated with GalNAc were provided by Alnylam Pharmaceuticals, Inc. GalNAc-siRNAs targeting MCJ were selected using in vitro cell-based activity screens in primary mouse hepatocytes with confirmation of activity in WT mice. The two selected siRNAs (si381 and si393) were conjugated to an N-Acetylgalactosamine (GalNAc) for optimized delivery in hepatocytes as previously described44. The siRNA duplexes targeting mouse MCJ (Dnajc15) were designed to maximize predicted efficacy and minimize off-target effects using proprietary algorithms (Alnylam). For in vivo screening, mice received a 10 mg/Kg s.c. dose in PBS of different MCJ-specific GalNAc-siRNA selected from the in vitro screening. After 2 weeks, liver expression of DnaJC15 was measured by Taqman RT-PCR using the following assay from ThermoFisher (Mm00481271_m1). GalNAc-si381 or GalNAc-si393 were selected from the in vivo studies and used for treatment of mice on the different diets at 10 mg/Kg in PBS administered s.c. every 2 weeks. Fully chemically-modified GalNAc-siRNA conjugates were employed to confer drug-like properties (Pharmacokinetics, pharmacodynamics, safety). Pattern of 2′-fluoro and 2′-O-methyl modifications were employed to improve in vivo potency45. si381 and si393 antisense strands target the following nucleotide sequence of the mouse MCJ transcript (NM_025384.3), respectively (404–424 and 358–378).
Healthy human liver (n = 5) samples from Valdecilla Hospital (Santander, Spain) were used as controls. All had histologically healthy liver, BMI <25 Kg/m2, normal fasting glucose, cholesterol and TG, normal AST and ALT, and no evidence of viral infections (HBV, HCV, and HIV). Diseased liver samples come from 21 patients with a clinical diagnosis of NAFLD who underwent a liver biopsy with diagnostic purposes in Santa Cristina Hospital (Madrid, Spain). Inclusion criteria for NAFLD patients were based on an alcohol intake lesser than 20 g/day, the presence of biopsy-proven steatosis with/without necroinflammation and/or fibrosis, and no evidence of hepatitis B and/or C virus (HBV and/or HCV, respectively) infection as well as human immunodeficiency virus (HIV) infection. The characteristics of the study groups are described in Table 1. Clinical examination included a detailed interview with special emphasis on both alcohol intake and medications use, history of known diabetes and arterial hypertension, as well as measurements of weight, height, blood pressure, and waist and hip perimeters. Body mass index (BMI) was calculated as weight (Kg) divided by height (m) squared. Fasting blood samples were obtained and used to measure alanine and aspartate transaminases (ALT and AST, respectively), GGT, total cholesterol, HDL-cholesterol, triglyceride, glucose, HbA1c, and insulin. Insulin resistance was calculated using the homeostatic model assessment (HOMA-IR) index1. Hepatic histopathological analysis was performed according to the scoring system of Kleiner et al.2. Four histopathological features were semi-quantitatively evaluated: grade of steatosis (0, <5%; 1, 5%–30%; 2, >30–60%; 3, >60%), lobular inflammation (0, no inflammatory foci; 1, <2 inflammatory foci per 200x field; 2, 2–4 inflammatory foci per 200x field; 3, >4 inflammatory foci per 200x field), hepatocellular ballooning (0, none; 1, few balloon cells; 2, many cells/prominent ballooning), and stage of fibrosis (from 0, none to 4, cirrhosis). Simple steatosis was defined as the presence of at least 5% of steatotic hepatocytes with or without mild lobular or portal inflammation but in the absence of features of hepatocellular injury (ballooning, apoptosis, or necrosis) and fibrosis. On the other hand, minimal criteria for the histological diagnosis of definite NASH included the combined presence of grade 1 steatosis, hepatocellular injury, and lobular inflammation with or without fibrosis. The study was performed in agreement with the Declaration of Helsinki, and with local and national laws. The Human Ethics Committee of Santa Cristina Hospital, the Comité Consultatif de Protection des Personnes dans la Recherche Biomédicale de Nice or the Human Ethics Committee of Valdecilla Hospital approved the study procedures, and written informed consent was obtained from all patients before inclusion in the study.
To perform H&E and Picro Sirius Red staining, paraffin-embedded liver samples were sectioned, dewaxed with a xylene substitute (HS-202, Histoclear; National Diagnostics, Atlanta, GA), and hydrated. Sections were stained for 5 min with Harry’s hematoxylin (HHS128-4L; Sigma) and for 15 min with aqueous eosin (HT110232-1L; Sigma) for H&E staining or with 0.01% Fast green FCF in saturated picric acid for 15 min and 0.04% Fast green For Coloring Food/0.1% Sirius red in saturated picric acid for 15 min for Sirius red staining. Samples were dehydrated and cleared with Histoclear. Finally, sections were mounted in DPX mounting media (06522, 500 mL; Sigma).
To perform Sudan III staining ornithine carbamyl transferase–embedded frozen samples were sectioned, cleared with 60% isopropanol, and stained with Sudan III solution (0.5% in isopropanol Sudan III Panreac) for 1 h and finally cleared with 60% isopropanol. Sections were counterstained with Mayer hematoxylin (MHS32-1L; Sigma) and mounted in aqueous mounting medium for lipid quantification.
To perform Oil Red O staining tissue sections from frozen tissue were stained with 0.2% Oil Red O and counterstained with hematoxylin.
To perform immunostaining in liver section, F4/80 samples were unmasked with proteinase K during 15 min at room temperature and MCJ samples with citrate buffer pH 6.0 during 20 min at 97 °C. Endogenous peroxidase activity was blocked for 10 min with 3% hydrogen peroxide, then sections were blocked with 5% normal goat serum for 30 min and incubated with F4/80 (1:50, 1 h at 37 °C, MCA497BB; Bio-Rad, Hercules, CA) or MCJ (B0027R BioMosaics) (1.100, overnight at 4 °C) followed by 30 min with anti-rat or anti-mouse reagents. Colorimetric detection was completed with Vector Vip purple substrate (sk-4600; Vector). Slides were counterstained with Mayer Hematoxylin (MHS32-1L; Sigma), and finally samples were dehydrated, cleared, and mounted in DPX mounting media (06522-500 mL; Sigma).
For the analysis, images from liver sections stained H&E, Picro Sirius Red, Sudam III, or Oil Red Oil were taken with an upright light microscope (Zeiss, Germany), or slides were scanned using the Leica-Aperio Versa (University of Vermont Imaging Facility). The average sum of intensities and stained area percentage of each sample was calculated using FRIDA software (http://bui3.win.ad.jhu.edu/frida/, John Hopkins University). No manipulation of the histological microscopy images was performed.
De novo lipogenesis analysis in liver
De novo lipogenesis was performed as previously described3 with slight modifications. In brief, freshly isolated tissue slices (40 mg) were incubated in high glucose DMEM with insulin (150 nM) and [H3] Acetic acid 20 µCi/ml for 4 h. Tissue slices were washed five times in cold PBS, homogenized in PBS and lipids were extracted and separated by TLC4,5, each lipid was scraped and the radioactivity was measured in a scintillation counter.
Beta oxidation was assessed as follows6,7. Fresh liver pieces were homogenized using a Potter homogenizer (5 strokes) in cold buffer (25 mM Tris-HCl, 500 nM sucrose, 1 mM EDTA-Na2 pH 7,4) and sonicated for 10 s. Then, the homogenates were centrifuged at 500 × g for 10 min at 4 °C. Approximately 500 µg of protein from the homogenate supernatants was used for the assay in a volume of 200 µl. The reaction started by adding 400 µl of assay mixture containing 0.5 µCi/ml [1-14C] palmitic acid to the samples and was incubated for 1 h at 37 °C in microfuge tubes with a Whatman paper circle in the cap. The reaction was stopped by adding 300 µl of 3 M perchloric acid and 1 M NaOH was added to impregnate the Whatman cap. After 2 hours the Whatman caps were retired and the radioactivity associated was measured in a scintillation counter. The microfuge tubes were centrifugated at 21,000 × g for 10 min at 4 °C. 400 µl of the supernatants were collected and the radioactivity was counted in a scintillation counter. The supernatant contained the ASM and the Whatman caps captured the released CO2.
Serum TGA were measured using a commercially available kit Triglycerides Liquid Mono (Krotest Laboratorios). Serum ketone bodies were measured using a commercially available kit from Wako Chemicals (Richmond, VA). Serum albumin and urea were measured using BioVision’s Albumin Assay Kit and Urea Assay Kit II, respectively (Milpitas, CA). Aspartate aminotransferase (AST) and alanine aminotransferase (ALT) were measured by multiple-point rate reflectance spectrophotometry on a Orhto Vitros 5600 analyzer. Serum insulin levels were determined using the Insulin (Mouse) ELISA kit (BioVision).
Isolation and culture of primary hepatocytes
Primary hepatocytes were isolated from male C57BL/6 WT mice via collagenase perfusion8. For in vitro silencing, WT primary hepatocytes were transfected with 100 nM MCJ siRNA using Jetprime reagent (Polyplus). Controls were transfected with an unrelated siRNA (Qiagen). For BODIPY staining, primary hepatocytes were incubated with 400 μM oleic acid and the complex I inhibitor Rotenone (Sigma-Aldrich) (1 μM) for the indicated times. Hepatocytes in culture were incubated with BODIPY 493/503 (Molecular Probes) at 1 mg/ml during 30 min prior to fixation (4% paraformaldehyde). Imaging was performed with an Axioimager D1 microscope and quantification of lipid bodies was performed using Frida Software.
Western blot analysis
Total protein and mitochondrial fraction protein from primary hepatocytes and liver tissue were resolved on sodium dodecyl sulfate-polyacrylamide gels and transferred to nitrocellulose or polyvinylidene difluoride (PVDF) membranes. Membranes were incubated with anti-mouse MCJ9 or antimouse GAPDH (Santa Cruz, Abcam) overnight. Primary antibodies were detected with anti-rabbit-IgG-HRP (Cell Signaling, Jackson Laboratory) and anti-mouse IgG-HRP (Santa Cruz Biotechnology) and signal developed with chemiluminescence (KPL) on an Amersham Imager 600.
RNA isolation and quantitative real-time polymerase chain reaction (RT-PCR)
Total RNA was isolated with Trizol (Invitrogen). Total RNA (1–2 μg) was treated with DNase (Invitrogen) and reverse transcribed into cDNA using M-MLV Reverse Transcriptase (Invitrogen). Quantitative real -time PCR (RT-PCR) was performed using SYBR® Select Master Mix (Applied Biosystems) and the Viia 7 Real-Time RT-PCR System (Applied Biosystems). The Ct values were extrapolated to a standard curve, and data was then normalized to the house-keeping expression (GAPDH). For real time RT-PCR of Acta2, we used the assay on demand primers and probe #Mm00725412 from ThermoFisher. Primers used for the rest of the other analyses are shown in Supplementary Table 1.
Mitochondrial respiration analysis
Liver mitochondrial respiration was measured at 37 °C by high-resolution respirometry using the Seahorse Bioscience XF24-3 Extracellular Flux Analyzer. For the measurement of the OCR, as the rate change of dissolved O2, liver mitochondria were isolated and plated in a XF24 cell culture microplate (Seahorse Bioscience), 5 μg per well as indicated10. Mitochondria were incubated in a media containing substrates for both complexes I and II: Glutamate (10 mM), Malate (2 mM), and Succinate (10 mM). After an OCR baseline measurement, sequential injections through ports in the XF Assay cartridges of pharmacologic inhibitors: ADP (4 mM) (state 3 respiration), Oligomycin (3 μM), an inhibitor of ATP synthase, which allows a measurement of ATP-coupled oxygen consumption through oxidative phosphorylation (OXPHOS); carbonyl cyanide 4-trifluoromethoxy-phenylhydrazone (FCCP) (4 μM), an uncoupling agent that allows maximum electron transport, and therefore a measurement of maximum OXPHOS respiration capacity; and finally Antimycin A (4 μM) + Rotenone (2 μM), mitochondrial complex I and III inhibitors respectively, were performed and changes in OCR were analyzed. The normalized data were expressed as pmol of O2 per minute or milli-pH units (mpH) per minute, per viability measured by MTT assay.
MDH2 activity was measured in liver extracts using the recommendation from the kit from Abcam ab119693.
ATP level determination
The levels of ATP in liver mitochondria were determined using the ATPlite luminescence ATP detection assay system (PerkinElmer) by following the recommendations from the manufacturer. Mitochondrial fractions were obtained using the Mitochondrial Fractionation Kit (ActiveMotif) for hepatic tissue.
ROS production analysis
For GSH quantification, liver extracts were analyzed with a UPLC system (Acquity, Waters, Manchester) coupled to a Time of Flight mass spectrometer (ToF MS, SYNAPT G2, Waters). A 2.1 × 100 mm, 1.7 μm BEH amide column (Waters), thermostated at 40 °C, was used to separate the analytes before entering the MS. Solvent A (aqueous phase) consisted of 99.5% water, 0.5% formic acid, and 20 mM ammonium formate while solvent B (organic phase) consisted of 29.5% water, 70% MeCN, 0.5% formic acid, and 1 mM ammonium formate. The extracted ion trace was obtained for GSH (m/z = 308.0916) in a 20 mDa window and subsequently smoothed (2 points, 2 iterations) and integrated with QuanLynx software (Waters, Manchester). Concentrations were converted into amount of analyte per mg liver tissue.
To evaluate lipid peroxidation in liver as marker of oxidative stress21,22, liver sections were immunostained for 4-HNE (Abcam, UK) as recommended by the manufacturer (Abcam). Immunostained images were visualized on an Axioimager D1 microscope, and the results were analyzed using the Frida software.
Statistical significance was determined by one-way analysis of variance or Student’s t test analysis, otherwise specified. All error bars represent standard deviation (SD), except when specified.
Comparison of DNAJC15 promoter CpG methylation with gene expression. DNA methylation IDAT files for 50 TCGA healthy liver samples46 were downloaded from the Genomic Data Commons (gdc.cancer.gov) and pre-processed using in-house pipelines. The pre-processing/filtering procedure left 451,561 CpGs in 47 noncancer liver tissues. The data set was further restricted to 40 subjects with matched RNA-seq data. The relation of the 3 DNAJC15 CpGs (cg14729962, cg09677945, cg15988970) and gene expression was determined by linear regression implemented in R version 3.5.1.
Comparison of DNAJC15 methylation in NAFLD to normal human liver. MCJ/DnaJC15 DNA methylation profiles from NAFLD patients and healthy human liver were publicly available in six GEO, (www.ncbi.nlm.nih.gov/geo) data sets listed in Table 2.
Two-tailed Wilcoxon’s rank sum tests were used to compare methylation of the three CpGs of interest (cg14729962, cg09677945, cg15988970)15 between liver tissue of NAFLD and healthy subjects. The total sample sizes for statistical comparisons were: 193, 190, and 141 for each of the above CpGs listed; cg09677945 was missing in GSE69852 (n = 3 healthy), and cg15988970 was missing in GSE61278 (n = 52 healthy). All statistical analysis and data visualization were performed in R version 3.5.1.
Further information on research design is available in the Nature Research Reporting Summary linked to this article.