Health benefits of late-onset metformin treatment every other week in mice

Chronic 1% metformin treatment is nephrotoxic in mice, but this dose may nonetheless confer health benefits if given intermittently rather than continuously. Here, we examined the effects of 1% metformin given every-other week (EOW) or two consecutive weeks per month (2WM) on survival of 2-year-old male mice fed standard chow. EOW and 2WM mice had comparable life span compared with control mice. A significant reduction in body weight within the first few weeks of metformin treatment was observed without impact on food consumption and energy expenditure. Moreover, there were differences in the action of metformin on metabolic markers between the EOW and 2WM groups, with EOW metformin conferring greater benefits. Age-associated kidney lesions became more pronounced with metformin, although without pathological consequences. In the liver, metformin treatment led to an overall reduction in steatosis and was accompanied by distinct transcriptomic and metabolomic signatures in response to EOW versus 2WM regimens. Thus, the absence of adverse outcomes associated with chronic, intermittent use of 1% metformin in old mice has clinical translatability into the biology of aging in humans.


Supplementary Information
. Physiological data. Effect of metformin on physical performance, metabolically relevant biomarkers, and hepatic hydrogen sulfide production in mice. Figure S2. Liver microarray data and AMPK activation levels at steady state.

Supplementary Tables ………………………………………………………………………………. 5-15
Table S1. Survival statistics for study mice stratified by metformin treatment groups. Table S2. Gross histopathology analysis of tissues collected at death or euthanasia. Table S3. List of the top 50 significantly enriched genes shared in the livers of EOW-SD and 2WM-SD pairwise comparisons.   Table S4c. List of the 45 significantly enriched GO Terms shared in the livers of EOW-SD and 2WM-SD pairwise comparisons. Table S5a. List of significant genes shared in the livers of 2WM-SD and CR40-AL pairwise comparisons. Table S5b. List of significant genes shared in the livers of EOW-SD and CR40-AL pairwise comparisons. Table S5c. List of significant genes shared in the livers of EOW-SD, 2WM-SD, and CR40-AL pairwise comparisons. Table S6a. List of significant liver metabolites in the EOW-SD and 2WM-SD pairwise comparisons. Table S6b. List of significant serum metabolites in the EOW-SD and 2WM-SD pairwise comparisons.    157.0 (-6) 163.6 (-3) The survival for EOW and 2WM groups is represented in weeks with the mean percentage change in survival relative to the SD group indicated in the parentheses. SD, standard diet; EOW, metformin treatment every-other-week; 2WM, metformin treatment for two consecutive weeks each month.  The zratios of significantly up-(red font) and down-(blue font) regulated genes in each pairwise comparison are shown. Significance is defined as zratio > 1.5 in either direction, false discovery rate (fdr) < 0.3 and p < 0.05. N = 4 mice per experimental group. EOW, metformin treatment every-other-week; 2WM, metformin treatment for two consecutive weeks each month; SD, standard diet.  The Z-scores of significantly up-(red font) and down-(blue font) regulated GO Terms in a given pairwise comparison are shown. Significance is defined as Z-score > 1.5 in either direction, false discovery rate (fdr) < 0.3 and p < 0.05. N = 4 mice per experimental group. EOW, metformin treatment every-otherweek; 2WM, metformin treatment for two consecutive weeks each month; SD, standard diet; n, number of genes in a given GO Term.
** GO Terms that follow opposite direction between EOW-SD and 2WM-SD pairwise comparisons.  The fold change of significantly up-(red font) and down-(blue font) regulated metabolites in each pairwise comparison is shown.

Supplementary Methods
Physical performance tests. Latency to fall from a rotarod, sole wire-hang and wire cage top and forelimb grip strength were measured in mice after 16 to 18 weeks on diet. Rotarod: mice were given a habituation trial where they were placed on a rotating rod (Med Associates, Inc., St. Albans, VT) at a constant speed of 4 r.p.m. and had to remain on it for 1 min before the accelerating rod trial. Time to fall from a rod that accelerates from 4 to 40 r.p.m. over a 5-min period was recorded and averaged over three trials on the same day. There was a 30-min rest period between each trial. Wire hang test: the animal was left hanged in a 3mm thick wire for a maximum of 1 min. Three attempts were recorded the same day and times were averaged. There was a 30-min rest period between each trial. Wire cage top: The animal was placed on a wire lid of a conventional housing cage and the lid was then inverted. The latency to when the animal falls was recorded and three attempts on the same day were averaged. The maximum trial length was 1 min and there was a 30-min rest period between each trial. Grip strength: The forelimb grasping applied by the mouse on a grid that is connected to a sensor (Columbus Instruments, Columbus, OH) was recorded. Five trials were carried out rejecting the highest and lowest values and averaging the other three.

Tissue collection.
A midline incision was made in the abdomen of anesthetized mice and the liver exposed. Blood was collected from the inferior vena cava and immediately after the liver was perfused in situ with Krebs-Henseleit buffer at low pressure to wash blood out of the sinusoids. Following this, liver, kidneys, white adipose tissue, brown adipose tissue and heart were excised and weighed. Liver portions were saved for biochemistry, histochemistry, scanning electron microscopy, microarray and metabolomics analyses.
Histology. At sacrifice, representative tissue sections from the liver and kidneys were fixed in 10% neutral buffered formalin (Electron Microscopy Sciences, Hatfield PA) and then embedded in paraffin. Sections of 10 µm were cut and mounted onto slides. Sections were stained with Hematoxylin and Eosin (H&E) and Periodic acid Schiff (PAS) stain according to standardized protocols. Quantification of hepatic steatosis in H&E sections and PAS staining was performed by three independent observers blinded to the treatment groups. This was performed on at least 10 fields per animal with the data representing the average of two scores. Steatosis was determined by assigning an equally weighted arbitrary score of 0 (no visible steatosis), 1 (< 33% of the section showing steatosis), 2 (>33% but < 66% of the section showing steatosis) to 3 (>66% of the section showing steatosis). PAS staining was determined by assigning an equally weighted arbitrary score of 0 (no visible PAS staining), 1 (< 33% of the section showing PAS staining), 2 (>33% but < 66% of the section showing PAS staining) to 3 (>66% of the section showing PAS staining).
Kidney sections from all groups of mice (SD, n=10; EOW, n=6; 2WM, n=6) were examined and scored for possible toxic lesions by a certified veterinarian pathologist. Both H&E and PAS stained slides were evaluated (see above). For tubulointerstitial lesions, the severity score was as followed: 1 -mild (patches of tubules with smaller basophilic renal epithelial cells (RTE) and no distortion of the renal capsule); 2-moderate (more obvious atrophy and attenuation of RTE. Basement membrane thickening. Foci take up about ½ of a cluster of proximal tubules. Mild distortion of capsular surface); 3-severe (patches of damaged tubules take up entire groups of proximal tubules. Most of the tubules in the patch have severe attenuation of RTE). In addition, the presence or absence of necrotic RTE, swollen RTE, basement membrane pigmentation, intraluminal casts, and inflammation were noted. For glomerular lesions, the severity score was as followed: 1 -mild (focal to multifocal segmental mesanglial matrix expansion); 2 -moderate (global mesanglial matrix expansion resulting in lobulation of glomerular tuft); 3 -severe (synechia, crescent formation and obsolescence). The presence of hyaline material in glomerular tufts (hyaline thrombi) was also noted. With regard to the distribution of lesions, a score 0 indicated no lesion; 1, up to 1/3 of tubules or glomeruli affected; 2, 1/3 -2/3 affected; and 3, >2/3 affected.
Electron microscopy. Livers were perfused with erythrocyte-free oxygenated Krebs-Henseleit bicarbonate buffer (95% O 2 -5% CO 2 , 37°C) until clear of blood, then a portion of the liver was subjected to needle perfusion (26G 1mL syringe) with fixative (2% glutaraldehyde-3% paraformaldehyde in 0.1 M sodium cacodylate buffer (0.1 M sucrose, 2 mM CaCl 2 ). Three to five randomly selected blocks from each animal were then dried and stained, mounted onto stubs, and sputter coated with platinum, as described previously 75 . A Jeol 6380 Scanning Electron Microscope (JEOL Ltd, Tokyo, Japan) at a resolution of 15,000× was used to examine fenestrations in the liver sinusoidal endothelium. Endothelial porosity is the percentage of the endothelial surface perforated by fenestrations and is calculated from their diameter and frequency.
Microarray method. Total RNA quantity and quality was tested using the Agilent Bioanalyzer RNA 6000 Chip (Agilent, Santa Clara, CA). Five hundred ng total RNA was labeled according to the manufacturer's instructions using the Illumina® TotalPrepTM RNA amplification kit (Illumina, San Diego, CA). A total of 750 ng biotinylated aRNA was hybridized to the Illumina Mouse Ref-8 v2 BeadChip overnight. Following posthybridization rinses, arrays were incubated with streptavidin-conjugated Cy3, and scanned at a resolution of 0.53 µm using an Illumina iScan scanner. Hybridization intensity data were extracted from the scanned images using Illumina BeadStudio GenomeStudio software, V2011.1. Raw data were subjected to Z-normalization, as described elsewhere 73,76 . Principal component analysis (PCA) was performed on the normalized Z-scores of all of the detectable probes in the samples using DIANE 6.0 software, available from: (http://www.grc.nia.nih.gov/branches/rrb/dna/ diane_software.pdf). Significant genes were selected by the z-test < 0.05, false discovery rate < 0.30, as well as z-ratio > 1.5 in both directions and ANOVA p value < 0.05.
For the calculation of pairwise distances between samples, each microarray was considered as a point in a high-dimensional space since we treated each probe as a variable. For parametric analysis of gene set enrichment (PAGE), our expression data was tested using the PAGE method as previously described 77 . Briefly, for each pathway under each pair of conditions, an aggregated Z score was computed as: Where pathway is the number of genes in the specific pathway and sample is the standard derivation of Zratio on the comparison sample arrays. For each Z (pathway) a P value was computed (JMP 6.0 software) to the total Z-ratio in comparison by Z-test. Ingenuity Pathways Analysis© was performed by using the tools supplied by Ingenuity Inc. (Ingenuity Systems; Redwood City, CA).
Liver microarray data from our recent study 32 on the effect of chronic 40% calorie restriction in 2-year-old male C57BL/6J mice versus ad libitum (AL) feeding (accession number GEO: GSE81959) was extracted and used to show the number of genes that overlapped with the intermittent metformin treatment (EOW-SD; 2WM-SD). Both control groups were comparable, as these mice (AL and SD) were of the same strain and age, and fed standard diet ad libitum. In the two studies, the Illumina Mouse Ref-8 v2 BeadChip was used, and Z-normalization of the raw data was carried out, enabling the selection of significant genes by the use of z-test, false discovery rate, z-ratio, and ANOVA. This exploratory data analysis was aimed only at visualizing trends in gene expression, using Venn diagrams, heatmap representation, and pie charts.
Quantitative RT-PCR. Total RNA was extracted from frozen tissue with Trizol® Reagent (Thermo Fisher Scientific, Waltham, MA). Briefly, 50-100mg of liver was homogenized in 1 mL Trizol® Reagent and phase separation was carried out with 0.2mL of chloroform. Aqueous phase is collected and RNA is precipitated with 100% isopropanol. Pellet is washed twice with 75% ethanol and let it dry. RNAis resuspended in RNAse free water and extracted quantity is measured with a NanoDrop 1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA). 1µg of RNA was reverse-transcribed using iScript™ cDNA Synthesis Kit ( Western blotting. Frozen liver tissues were lysed in radioimmunoprecipitation buffer containing EGTA and EDTA (Boston BioProducts, Ashland, MA, USA) supplemented with protease inhibitor cocktail (Sigma-Aldrich) and phosphatase inhibitor cocktail sets I and II (Calbiochem, San Diego, CA, USA). Protein concentration in clarified lysates was determined using the bicinchoninic acid reagent (Pierce BCA Protein Assay Kit, Thermo Fisher Scientific, Waltham, MA, USA). Proteins (20 µg/well) were separated on 4-15% Criterion TGX precast gels (BioRad, Hercules, CA, USA) using SDS-polyacrylamide gel electrophoresis under reducing conditions and then electrophoretically transferred onto nitrocellulose membranes (Trans-Blot Turbo Transfer System, BioRad). Western blots were performed according to standard methods, which involved a blocking step in phosphate-buffered saline/0.1% Tween-20 (PBS-T) supplemented with 5% non-fat milk and incubation with primary antibodies of interest. All antibodies were detected with horseradish peroxidase-conjugated secondary antibodies (Santa Cruz Biotechnology, Dallas, TX, USA) and visualized by enhanced chemiluminescence (Immobilon Western Chemiluminescent HRP Substrate, Millipore, Billerica, MA, USA). Imaging of the signal was captured with Amersham Imager 600 (GE Healthcare, Piscataway, NJ, USA). Quantification of the protein bands was performed by volume densitometry using ImageJ software (National Institutes of Health, Bethesda, MD, USA). The primary antibodies used in this study were raised against AMPK (cat.: 2532S, Cell Signaling Technology, Danvers, MA, USA), phosphoThr-172 AMPK (cat.: 2531S, Cell Signaling), ACC (cat.: 04-322, Millipore), and phosphoSer-79 ACC (cat.: 07-303, Millipore). The antibodies were used at the dilution (1:1000) recommended by the manufacturers.
Metabolomics. Liver tissue (4mg) was homogenized in extraction solution by mixing acetonitrile, isopropanol and water in proportions 3:3:2 (JT Baker, Center Valley PA), then vortexed for 45 s and then 5 min at 4C. Following centrifugation for 2 min at 14,000 rcf, two aliquots of the supernatant (500µL each aliquot) were made for analysis and one for backup. One aliquot was dried via evaporation overnight in the Labconco Centrivap cold trap concentrator (Labconco, Kansas City MO). The dried aliquot was then re-suspended with 500µL 50% acetonitrile (degassed as given), then centrifuged for 2 min at 14,000 rcf using the centrifuge Eppendorf 5415. The supernatant was moved to a new Eppendorf tube and again evaporated to dryness. Internal standards (C08-C30, fatty acid methyl esters) were then added and the sample was derivatized by methoxyamine hydrochloride in pyridine and subsequently by N-methyl-Ntrimethylsilyltrifluoroacetamide for trimethylsilylation of acidic protons. Data were acquired using the method as described in (Fiehn et al., 2008). Briefly, metabolites were measured using a Restek corporation rtx5Sil-MS column (Restek Corporation; Bellefonte PA; 30 m length x 0.25 mm internal diameter with 0.25µm film made of 95% dimethly/5%diphenylpolysiloxane) protected by a 10m long empty guard column which is cut by 20cm intervals whenever the reference mixture QC samples indicate problems caused by column contaminations. This sequence of column cuts has been validated by UC Davis Metabolomics Core with no detrimental effects detected with respect to peak shapes, absolute or relative metabolite retention times or reproducibility of quantifications. This chromatography method yields excellent retention and separation of primary metabolite classes (amino acids, hydroxyl acids, carbohydrates, sugar acids, sterols, aromatics, nucleosides, amines and miscellaneous compounds) with arrow peak widths of 2-3s and very good within-series retention time reproducibility of better than 0.2s absolute deviation of retention times. The mobile phase consisted of helium, with a flow rate of 1 mL/min, and injection volume of 0.5µL. The following mass spectrometry parameters were used: a Leco Pegasus IV mass spectrometer with unit mass resolution at 17 spectra s-1 from 80-500 Da at -70 eV for elution of metabolites. As a quality control, for each sequence of sample extractions, one blank negative control was performed by applying the total procedure (i.e. all materials and plastic ware) without biological sample. Result files were transformed by calculating the sum intensities of all structurally identified compounds for each sample (i.e. those signals that had been positively identified in the data pre-processing schema outlined above), and subsequently dividing all data associated with a sample by the corresponding metabolite sum. The resulting data were multiplied by a constant factor in order to obtain values without decimal places. Intensities of identified metabolites with more than one peak (e.g. for the syn-and anti-forms of methoximated reducing sugars) were summed to only one value in the transformed data set. The original non-transformed data set was retained. The general concept of this data transformation is to normalize data to the 'total metabolite content', but disregarding unknowns that might potentially comprise artifact peaks or chemical contaminants.
Sample preparation for analysis and quantification of metformin. 1) Mouse serum analysis. To 20 µl serum, 5 µl internal standard (IS) solution (3200 ng/ml phenformin (Sigma-Adrich) in methanol), 5 µl calibration standards (40-20000 ng/ml), QCs, and study samples. For blanks, 5 µl of methanol were added to the blank mouse serum in place of the IS solution. 2) Mouse liver samples. Liver samples were stored frozen at -70°C and thawed to room temperature, then homogenized in phosphate-buffered saline (PBS) using a probe sonicator (30 s at 50% maximal amplitude), using 4 ml PBS per gram liver tissue (ml/g). To 100 µl liver homogenates, 20 µl IS solution, 20 µl ammonium hydroxide solution (15% in water), and 400 µl acetonitrile were added to the standards, QCs, and study samples. For blanks, 20 µl of methanol were added to the blank mouse serum in place of the IS solution. For both serum and liver samples, these mixtures were vortexed for 10 min on a multi-tube vortex mixer at maximal speed, and the suspensions were then clarified by centrifugation (18000 g, 10 min). Five (serum) and eight (liver) microliters of the resulting supernatants were added to 1000 µl of 90% methanol in water, and then these mixtures were transferred to HPLC vials for LC-MS/MS analysis as described below. Accuracy and precision during sample analysis for metformin in mouse serum and liver homogenates were as followed: serum (94.0% accuracy, and CV of 10.1%) and liver homogenate (100% accuracy, and CV of 9.17%).