A total of 48 C57Bl/6 mice (age, 8–12 weeks old; weight, 18–20 g) were purchased from Shanghai Slack Experimental Center (Shanghai, China) and were housed in a specific pathogen free (SPF) environment (24–26°C; relative humidity 50–60%) with a 12-h light/dark cycle and free access to food and water. Dulbecco's modified Eagle's medium (DMEM) and Hank's balanced salt solution (HBSS) were obtained from Gibco (Thermo Fisher Scientific, Inc., Waltham, MA, USA). Hepatocyte medium was purchased from ScienCell Research Laboratories, Inc. (Carlsbad, CA, USA). Sodium pyruvate, sodium lactate, dexamethasone, bovine serum albumin (BSA), 8-bromoadenosine 3′,5′-cyclic monophosphate (8-br-cAMP) and metformin were acquired from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany). All the primers used in RT-qPCR were synthesized by Shanghai Sangon Biological Engineering Technology and Services Co., Ltd. (Shanghai, China).

All experiments were supervised and approved by the laboratory ethics committee of Ruijin Hospital affiliated with Shanghai Jiaotong University School of Medicine (Shanghai, China). Hepatocytes were isolated from 8–12 week old male C57Bl/6 mice. Briefly, following anesthesia, mouse livers were perfused with 10 ml 1X HBSS without calcium, followed by perfusion with 0.05% collagenase IV in calcium-containing HBSS in a recirculating manner. The liver was then detached and filtered through a 70 μm nylon mesh and hepatocytes were sedimented by centrifugation at 200 × g for 5 min at 4°C. The hepatocytes were plated onto 6-well plates and grown in hepatocyte medium supplemented with BSA, penicillin/streptomycin and hepatocyte growth factor. The cultures were maintained at 37°C in humidified atmosphere (95% air and 5% CO₂). After 24 h, the cells had become attached to the plates and the medium was replaced with low glucose DMEM containing 0.25% BSA with dexamethasone (100 nM) for prestimulation for 12–16 h. The cells were then incubated with 100 μM 8-br-cAMP (a permeable analogue of cAMP) in the presence or absence of 2 mM metformin in glucose-free DMEM containing gluconeogenic substrates (10 mM sodium lactate and 1 mM sodium pyruvate) for 8 h.

Total RNA was extracted from primary mice hepatocytes using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. The RNA purity and concentration was evaluated with a NanoDrop-2000 spectrophotometer (Thermo Fisher Scientific, Inc.). The optical density (OD) absorbance ratio at 260 and 280 nm (A₂₆₀/A₂₈₀ ratio) was ~2.0 and the OD A₂₆₀/A₂₃₀ ratio was ~1.8. RNA integrity was determined by standard denaturing agarose gel electrophoresis, which revealed that the samples were good quality.

ArrayStar mouse LncRNA microarray version 2.0 (ArrayStar Inc., Rockville, MD, USA) is designed for the global profiling of mouse lncRNAs in hepatocytes isolated from 3 mice (i.e., 3 groups at 3 replicates). The sample preparations and microarray hybridization were performed based on the manufacturer's protocols. Briefly, 1 μg total RNA from each sample was amplified and transcribed into fluorescent cRNA using Agilent's Quick Amp Labeling kits (Agilent Technologies, Inc., Santa Clara, CA, USA) according to the manufacturer's protocol. The labeled cRNAs were hybridized onto the mouse 8×60K LncRNA array. After washing, the arrays were scanned using the Agilent G2505C Microarray Scanner System, and Agilent Feature Extraction software was used to analyze the acquired array images (Agilent Technologies, Inc.). Quantile normalization and subsequent data processing were performed using the GeneSpring GX v12.0 software package (Agilent Technologies, Inc.). The threshold set for significantly differential genes was an absolute log value of fold-change ≥2 and a P-value ≤0.05. The microarray analysis was performed by KangChen Bio-tech Inc. Corporations (Shanghai, China).

Following total RNA extraction, as described above, cDNA was synthesized by reverse transcription using Superscript II RT reagent kit (Promega Corporation, Madison, WI, USA) according to the manufacturer's protocol. RT-qPCR was performed with the Roche LightCycle^r® 480 instrument (Roche Diagnostics, Basel, Switzerland) using SYBR Premix Ex Taq (Takara Bio, Inc., Otsu, Japan) in a final volume of 10 μl. The PCR conditions were as follows: Denaturation at 95°C for 10 sec, 40 cycles at 95°C for 5 sec and 60°C for 31 sec. A melting curve was constructed in the temperature range of 60 to 95°C at the end of the amplification. The sequences of the primers used are shown in Table I. Relative gene expression levels were quantified based on the cycle threshold (Cq) values and normalized to the internal control gene β-actin, which received a value of 1. The relative gene expression was calculated using the 2^−ΔΔCq method (16).

All values are presented as mean ± standard deviation from at least three independent experiments, and P≤0.05 was considered to indicate a statistically significant difference. Differential expression levels of genes were compared using ANOVA for multiple groups or Student's t-test for two groups. Statistical analysis was performed by SPSS 18.0 software (IBM, Armonk, NY, USA).

Gluconeogenesis is tightly controlled through the transcriptional regulation of PEPCK, the rate-limiting enzyme of hepatic gluconeogenesis (17). Metformin has been reported to suppress PEPCK expression in primary hepatocytes (14). Prior to microarray analysis, the effect of metformin on PEPCK mRNA expression in the primary mouse hepatocytes was investigated. Following the incubation of primary mouse hepatocytes with 100 μM 8-br-cAMP for 8 h, PEPCK mRNA expression was significantly increased, and this increase was significantly suppressed by 2 mM metformin (Fig. 1). This indicates that the experimental conditions are suitable for studying the mechanism of gluconeogenesis. Therefore, the microarray analysis of the lncRNAs expression profile was performed using the same conditions.

A total of 22,016 lncRNAs were detected in the primary mouse hepatocytes using ArrayStar Mouse LncRNA Microarray version 2.0. The lncRNAs represented in this microarray have been sourced from authoritative databases, including RefSeq, UCSC Knowngenes and Ensembls. Scatter plots were used to assess the variation in the expression of lncRNAs between the cAMP group and the control group (Fig. 2A) or between the metformin group and the cAMP group (Fig. 2B). By setting a threshold for differential expression at fold-change >2.0-fold, 865 differentially expressed lncRNAs (456 upregulated and 409 downregulated lncRNAs) were identified in the cAMP group vs. the control group. Compared with the cAMP group, 4,580 differentially expressed lncRNAs were identified, of which 2,114 lncRNAs were upregulated and 2,466 lncRNAs were downregulated in the metformin group. Among the cAMP-upregulated lncRNAs, 189 were downregulated by metformin. Among the cAMP-downregulated lncRNAs, 167 were upregulated by metformin. Heatmaps demonstrate the lncRNA expression profiles of the groups, and the expression values of the 30 most strongly cAMP-upregulated and cAMP-downregulated lncRNAs (Fig. 2C and D and Table II) reversed by metformin are presented.

It is likely that these metformin-reversed lncRNAs serve a critical role in the regulation of gluconeogenesis. Therefore, some general signatures of these lncRNAs, such as their sources, classification, length distribution and chromosome distribution, were investigated. A pie chart was constructed to show the number of metformin-reversed lncRNAs collected from different databases (Fig. 3A). There were 78 antisense overlap, 25 bidirectional, 214 intergenic and 39 sense overlap lncRNAs among these lncRNAs (Fig. 3B), which were mainly between 400 and 3,600 nt in length (Fig. 3C). Chromosome distribution analysis demonstrated that these metformin-reversed lncRNAs were located on a number of different chromosomes (Fig. 3D and E).

Among the identified metformin-reversed lncRNAs, 8 cAMP-stimulated and metformin-inhibited l ncR NAs (A K1336 02, ENSM UST 0 0 0 0 0138573, ENSMUST00000129953, uc009njr.1, AA591058, NR_027710, NR_030715 and ENSMUST00000145208) were selected for analysis by RT-qPCR in order to verify the microarray data. The expression levels of all 8 cAMP-stimulated lncRNAs were significantly decreased in the presence of metformin as shown by RT-qPCR (Fig. 4), consistent with the results of the microarray analysis.

Ponjavic et al (18) emphasized the importance of lncRNA and its adjacent protein-coding gene pairs when investigating the function of lncRNAs. Therefore, using the UCSC genome browser (http://genome.ucsc.edu/) and NONCODE database (http://www.noncode.org), the sequences of the eight validated lncRNAs and their associated coding genes were obtained (data not shown). The lncRNA ENSMUST00000138573 is a 614-nt antisense overlapping lncRNA associated with the G protein-coupled receptor 155 (Gpr155) gene. LncRNA NR_027710 is a sense overlapping lncRNA, which is located near the PGC-1α gene. Notably, RT-qPCR analysis demonstrated that Gpr155 and PGC-1α displayed a similar expression pattern to their associated lncRNAs under the same treatment conditions (Fig. 5A and B). Therefore, it is possible that the two lncRNAs modulate gluconeogenesis through their associated protein-coding genes.