Dulbecco’s Modified Eagle’s Medium low glucose (DMEM), l-glutamine, penicillin/streptomycin (10,000 IE/10 mg/ml), HEPES, amphotericin B and l-carnitine were from Sigma-Aldrich, St. Louis, MO, US. Foetal calf serum (FCS), Dulbecco’s Phosphate Buffered Saline (DPBS) and trypsin/EDTA (0.05%) were from Gibco/Invitrogen, Grand Island, NY, USA. Ultroser G was from Pall Corporation, St-Germain-en-Laye Cedex, France. D-[1-14C]glucose (54 mCi/mmol), [14C]deoxy-d-glucose (287 mCi/mmol) and [1-14C]oleic acid (53 mCi/mmol) were purchased from PerkinElmer, NEN, Boston, MA, USA. D-[14C(U)]glucose (5 mCi/mmol) was provided by American Radiolabeled Chemicals Inc., St. Louis, MO, USA. Insulin Actrapid® was from Novo Nordisk, Bagsvaerd, Denmark. Corning® CellBIND® microplates were from Corning B·V. Life Sciences, Schipol-Rijk, the Netherlands, 96-well UNIFILTER® microplate from Whatman, Middlesex, UK, and 96-well Isoplate, ScintiPlate®-96 TC micoplates, and the scintillation liquid Optiphase Supermix was from Perkin Elmer, Waltham, Massachusetts, USA. Bio-Rad Protein Assay Dye Reagent was from Bio-Rad Laboratories, NY, USA. Agilent Total RNA isolation kit was obtained from Agilent Technologies (Santa Clara, CA, USA). Illumina Human-6 Express BeadChips version 3 arrays were from Illumina (San Diego, CA, USA).
Human skeletal muscle cell cultures
Satellite cells were isolated from the M. obliquus internus abdominis of 6 healthy donors, age 38.8 (±3.8) years, body mass index 22.1 (±1.5) kg/m2, fasting glucose 5.0 (±0.1) mM, insulin, plasma lipids and blood pressure within normal range and no family history of diabetes. The biopsies were obtained with informed consent and approval by the Regional Committee for Research Ethics, Oslo, Norway. The cells were cultured in DMEM (5.5 mM glucose) with 2% FCS, 2% Ultroser G, l-glutamine (4 mM), penicillin/streptomycin (P/S) and amphotericin B until 70–80% confluent. Myoblast differentiation to myotubes was then induced by changing medium to DMEM (5.5 mM glucose) with 2% FCS, 25 pM insulin, l-glutamine (4 mM), P/S and amphotericin B. Experiments were performed after 8 days of differentiation, and preincubation with benfotiamine (100–200 μM) and/or hyperglycemia (20 mM glucose) was started after 4 days.
Substrate oxidation assay
The muscle cells were cultured on 96-well CellBIND® microplates. Substrate, [U-14C]glucose (1 μCi/ml, 200 μM) or [1-14C]oleic acid (1 μCi/ml, 100 μM) was given in DPBS with 10 mM HEPES (1 mM l-carnitine was also added with oleic acid as described previously (Aas et al. 2011)). A 96-well UNIFILTER® microplate was mounted on top of the CellBIND® plate as described before (Wensaas et al. 2007), and the cells were incubated at 37°C for 4 h. The CO2 trapped in the filter was counted by liquid scintillation (MicroBeta®, PerkinElmer). The remaining cell-associated radioactivity was also assessed by liquid scintillation, and the sum of CO2 and cell-associated radioactivity was considered as total substrate utilization. Protein content in each well was determined (Bradford 1976), and the data are presented as CO2/mg protein or cell-associated radioactivity/mg protein.
Scintillation proximity assay
Radiolabeled substrates taken up and accumulated by adherent cells were measured by scintillation proximity assay (SPA). The cells were grown and differentiated in 96-well ScintiPlate®-96 TC SPA plates. Measurements of [14C]deoxyglucose (1 μCi/ml) uptake by SPA were taken in DMEM without phenol red (Sigma, MO, cat. no. D5030) with additional 100 μM glucose. Finally, the cells were washed 3 times with PBS and harvested with 0.1 M NaOH (200 μl/well). Protein was determined according to Bradford et al. (Bradford 1976).
RNA isolation and microarray analysis
Total RNA was prepared from primary myotubes from three donors using Agilent Total RNA isolation kit according to the supplier’s protocol (Agilent Technologies). RNA was used individually, and RNA integrity was checked on chip analysis (Agilent 2100 bioanalyzer, Agilent Technologies, Santa Clara, CA, USA) according to the manufacturer’s instructions. RNA was judged as suitable for array hybridization only if samples exhibited intact bands corresponding to the 18S and 28S ribosomal RNA subunits and displayed no chromosomal peaks or RNA degradation products (RNA Integrity Number >9.0). cRNA synthesis was performed using Illumina TotalPrep RNA Amplification (San Diego, CA, USA) according to the supplier’s protocol. Hybridization, washing and scanning of Illumina Human-6 Express BeadChips version 3 arrays (San Diego, CA, USA) were according to standard Illumina protocols. Data extraction and quality control were performed using BeadStudio version 126.96.36.199 (Illumina) and the Gene Expression module 3.2.7. Arrays were normalized using quantile normalization, and expression estimates were calculated by GC robust multiarray average background adjustment. Gene set enrichment analysis (GSEA; Subramanian et al. 2005) was used to test the specific hypothesis that groups of genes involved in glucose metabolism were changed during benfotiamine treatment. Collections of gene sets were obtained using a gene set browser from the Broad Institute website http://www.broad.mit.edu/gsea/ (Molecular Signatures Database v3.0). This browser searches a number of publicly available sources (e.g., Biocarta, Kegg) where genes are grouped if they belong to the same pathway and share ontology terms or clinical phenotypes. Thus, it is possible to define specific gene sets on the basis of a particular parameter of interest. Similar use of GSEA has been published earlier (Bohn et al. 2010). Gene set collections relevant to glucose and energy metabolism were obtained using the following keywords stepwise: “glucose,” “insulin,” “energy,” “lactate,” “mitochondria,” “transporter*” (*indicates truncated search), “lipid,” “DNA and Repair” and “interleukin.” The larger, predefined gene set collections: C2, C3TFT and C5 from http://www.broad.mit.edu/gsea/ were also tested. The C3 TFT (transcription factor targets) collection consists of gene sets that contain genes sharing a transcription factor-binding site defined in the TRANSFAC (version 7.4, http://www.gene-regulation.com/). The gene sets in the C5 collection are grouped according to the gene ontology consortium. The C2 gene sets are collected from various sources such as online pathway databases, publications in PubMed and knowledge of domain experts. The benfotiamine-treated samples were compared with untreated in paired GSEA analysis. GSEA was performed using J-express 2011 (http://www.molmine.com) according to the description on the J-express manual. The gene matrix was collapsed by selecting the maximum probes. Log-fold change was used as the scoring method, the number of permutations was set to 1,000 and gene sets with less than 10 genes or more than 500 genes were excluded from the analysis. False discovery rate (FDR) q values <5% were used as criteria for significantly enriched gene sets. Minimum Information About a Microarray Experiment (MIAME) standards (Brazma et al. 2001) were followed in the analysis and storage of microarray data. The raw data are available at the gene expression omnibus (GEO) at http://www.ncbi.nlm.nih.gov/geo/by accession number GSE31553.
RNA isolation and analysis of gene expression by TaqMan® qPCR
Cells were harvested, and total RNA was isolated by Agilent Total RNA isolation kit (Agilent Technologies, Santa Clara, CA, USA) according to the supplier’s total RNA isolation protocol. Total RNA was reverse-transcribed with oligo primers using a Perkin-Elmer Thermal Cycler 9600 (25°C for 10 min, 37°C for 1 h 20 min and 85°C for 5 s) and a TaqMan reverse transcription reagent kit (High Capacity cDNA Reverse Transcription Kit, Applied Biosystems, USA). Two micrograms of total RNA was added per 20 μL of total TaqMan reaction solution. Real-time PCR was performed using an ABI PRISMT 7000 Detection System (Applied Biosystems, USA). RNA expression was determined by SYBRT Green, and primers were designed using Primer ExpressT (Applied Biosystems, USA). Each target gene was quantified in triplicate and carried out in a 25 μL reaction volume according to the supplier’s protocol. All assays were run for 40 cycles (95°C for 12 s followed by 60°C for 60 s). The housekeeping control genes GAPDH (FC −1.1, P = 0.8 from microarray) and 36B4 were both measured, and transcription levels are presented as averaged change relative to levels of GAPDH and 36B4. Primer sequences: GAPDH (acc.no. NM_008084): F: CATGGCCTTCCGTGTTCCT, R: TGATGTCATCATACTTGGCAGGTT; 36B4 (acc.no. NM_007475): F: ATCTCCAGAGGCACCATTGAA, R: TCGCTGGCTCCCACCTT; NOX4 (acc.no. NM_016931): F: TGGACCTTTGTGCCTGTACTGT; R: TGAGGATGACTTATGACCGAAA.
All data are presented as mean ± SEM. Statistical comparison between different treatments was performed by linear mixed model (LMM) and ANOVA repeated measures using SPSS ver. 17.0 (SPSS Inc., Chicago, IL, USA). The parameter of interest was entered as the dependent variable, and pretreatment (with and without benfotiamine) and acute treatments were entered as fixed variables. Differences were considered statistically significant at P values less than 0.05. All experiments were performed with at least triplicate observations, and replicate experiments were performed on cells from different donors. For microarray analysis, individual fold change (FC) for each donor was calculated by dividing the expression level after benfotiamine treatment on the expression level after control treatment. The individual fold change was then log2 transformed. Mean fold changes were calculated based on the log2-transformed individual fold changes, followed by the identification of differentially expressed probe sets using intensity-based moderated t-statistics (Sartor et al. 2006). P values were corrected for multiple testing using Benjamini and Hochberg’s false discovery rate (FDR) method (Benjamini and Hochberg 1995). Probe sets that satisfied the criterion of FDR <10% (q value <0.1) and fold change >1.5 or <−1.5 were considered to be significantly regulated.