Serum from humans on long-term calorie restriction enhances stress resistance in cell culture

Gene expression analysis shows up-regulation of genes involved in NRF2-mediated oxidative stress response

Mammalian fibroblasts have been extensively used to assess global transcriptional changes induced by circulating factors (i.e. serum) on cell function [9]. To test the hypothesis that the CR-induced changes in circulating factors can influence cell function, we compared the transcriptional profiles of human primary fibroblasts treated with CR or WD serum. Differentially expressed genes were selected using a moderate t-test approach [10] and multiple test correction [11], and these genes were used to identify transcriptionally altered pathways. Ingenuity Pathways Analysis software (IPA) showed a significant association of a subgroup of 83 genes (Supplementary Table S1) with many oxidative response-related pathways (Fig. 1). In particular, NF-E2-related factor 2 (Nrf2)-mediated oxidative stress response contained 30 genes that were dysregulated in the treated fibroblasts (BH p-value 3.33E-02), and many of them were direct expression effectors of Nrf2 activity (Fig. 2). Meta-analysis performed using a competitive gene set test [12] pinpoints 11 molecular signatures significantly down-regulated in a ‘self-contained’ gene test (Table 2 red Highlights). To further investigate the result of microarray data and confirm the association with oxidative response-related pathways, we performed quantitative real-time PCR for the following genes which were shown by microarray to be detectable and significantly dysregulated: GSTK1, GPX1, SOD2, GSTT1, and IDH2. As shown in Figure 3, relative abundance of message for these genes was significantly higher in the CR group than the control group (p<0.0005) with a fold change difference ≥ 2.2-fold.

Incubation with human CR serum increases stress resistance to hydrogen peroxide in vitro

To confirm the gene expression findings with a functional test, human primary fibroblasts cultured for 48 hours with CR or WD sera were exposed to increasing concentrations of the powerful oxidizing agent, hydrogen peroxide (H₂O₂) for 24 hours. As shown in Figure 4, treatment with H₂O₂ produced a dose-dependent decline in survival, but fibroblasts cultured with CR serum were significantly protected from the cytotoxic effect of the oxidant. At 600 μM about 67% of cells pretreated with CR serum were viable, while only 37% of cells pretreated with WD serum survived (p< 0.001), and even at higher H₂O₂ concentrations (800 μM) 32% of the cells pretreated with CR serum survived, while only 10% of the cells exposed to WD serum were still alive (p< 0.01).

Human subjects

Human serum samples were collected in the morning after an overnight fast from twelve individuals of the Calorie Restriction Society who had been practicing severe CR with adequate nutrition for an average of 7 years (range 3-20 years). Thirteen healthy, non-obese and sedentary (regular exercise < 1 h per week) individuals eating typical Western Diets (WD) were used as a sedentary comparison group; they were matched with the CR group in terms of age, sex and height. Whole blood was collected into sterile vacutainer tubes, allowed to clot for 30 minutes and then centrifuged. The serum was then stored at −80°C. The characteristics of the participants are shown in Table 1. All of the study participants were weight stable, i.e. < 2 kg weight change in the preceding 6 months. None of the participants had evidence of chronic disease (including cardiovascular, lung, gastrointestinal, type 2 diabetes and cancer). None were smokers. This study was approved by the Human Studies Committee of Washington University School of Medicine, and all participants gave informed consent before their participation.

Cell culture

The BJ human fibroblast cell line was obtained from ATCC and used in these experiments. BJ cells were cultured in 75 cm² culture flasks containing basic culture medium: 86% of E-MEM medium supplemented with 10% fetal bovine serum (Sigma-Aldrich, St. Louis, MO, USA), 1% Sodium Pyruvate, 1% non-essential amino acids, 1% GlutaMAX and 1% antibiotics (GIBCO, Carlsbad, CA, USA) until the time of treatment. Cells were cultured in a humidified incubator at 37 °C and 5% CO₂. For the experiments, FBS was replaced with human serum from CR and WD participants. Both FBS and human serum were heat inactivated for 30 minutes at 55 °C.

Microarray analysis

BJ cells were seeded at a concentration of 4.000 cells/cm² in 25 T-25 flasks; when they reached 60% of confluence, the cells were starved for 24h without FBS. After starvation, 10% of human serum from 12 CR and 13 WD individuals was added to the cultures for 48h. Whole cells were detached from the flask with trypsin/EDTA (Sigma-Aldrich, St. Louis, MO, USA) and Total RNA was extracted using the mirVANA (Ambion, Inc, Austin, TX, USA) according to manufacturer's protocol. Integrity of total RNA was evaluated using capillary electrophoresis (Bioanalyzer 2100, Agilent technologies, Santa Clara, CA, USA) and quantified using a Nanodrop 1000 (Nanodrop, Wilmington, DE). Aliquots of RNA (400 ng) samples were amplified according to the specifications of the Illumina® TotalPrep™ RNA Amplification Kit (Ambion, Austin, TX, USA) to produce a pool of biotin-labeled RNA corresponding to the polyadenylated (mRNA) fraction. The cRNA samples were applied to the arrays of Illumina whole-genome HumanHT-12 v 3.0 (Illumina, San Diego, CA, USA) and hybridized according to the manufacturer's specification. Each array on the BeadChip targets over 25.000 annotated genes using 3-micron beads bearing covalently attached 50-base oligonucleotide probes. Each probe interrogates a single gene, and each bead type is represented with an average 15-fold redundancy on every array. The BeadChips were scanned with the Illumina's Beadarray system scanner (Illumina, San Diego, CA, USA). The hybridization images signal intensity was extracted and background subtracted using Illumina Inc. BeadStudio software version 3.3.7. The produced data were checked for the Illumina internal quality control and loaded into Bioconductor software [23,24] for statistical analysis using lumi package [25] (see below). The data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus [26] and are accessible through GEO Series accession number GSE41790.(http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?token=nnonbsicqouiilw&acc=GSE41790).

Real Time PCR

Real-time PCR was used to verify microarray data. Total RNA was DNase I treated and reverse transcribed using the High Capacity RNA-to-cDNA kit (Applied Biosystems, Carlsbad, CA, USA). The cDNA was then used to evaluate the relative expression of select genes including SOD2, GPX1, IDH2, GSTK1 and GSTT1 using specific TaqMan Gene Expression Assays (Applied Biosystems, Carlsbad, CA, USA), which are pre-designed and pre-optimized gene-specific probe sets. DNA amplification was performed using the Applied Biosystems 7500 Fast Real-Time PCR system according to the manufacturer's protocol. The relative amount was calculated from the threshold cycles with the instrument's software (SDS 2.0, Applied Biosystems, Carlsbad, CA, USA), according to the manufacturer's instructions. We used the comparative cycle threshold (Ct) method (2 - CT) to calculate relative gene expression under experimental and control conditions normalized to GAPDH. The results are expressed either as the abundance of the gene of interest relative to the house keeping gene (2 - CT), or as fold change over control values (2 - CT).

Cell Proliferation assays in response to Hydrogen Peroxide

To determine the Differential Stress Resistance to hydrogen peroxide of serum from CR and WD individuals, human fibroblasts were seeded in 96-well plates at a concentration of 3000 cells/well in 86% D-MEM low glucose w/o phenol red supplemented with 1% Sodium Pyruvate, 1% non-essential amino acids, 1% GlutaMAX and 1% antibiotics (GIBCO, Carlsbad, CA, USA). After cells had seeded, media was removed and replaced with media containing 10% of human serum from 11 CR and 11 WD individuals was added in triplicate to the plates for 48hs. After incubation, freshly prepared hydrogen peroxide was added to the plates at a concentration of 200, 400, 600, 800, 1000 μM and cells were challenged for 24hs.

Cell viability was determined by the addition of 20 μl of the ready-to-use WST1 reagent (Roche diagnostics, Mannheim, Germany) to each well; the cells were incubated for approximately 3 hours at 37 °C and 5% CO₂ in a humidified incubator. After this incubation period, the absorbance of the formazan dye formed was measured at a wavelength of 450 nm using a microplate spectrometer. The measured absorbance directly correlates to the number of viable cells.

Statistical analysis

Microarray data produced were quality checked with arrayQualityMetrics [24] package under R version 2.15.1 Bioconductor software [23] and undetectable and low detection value probes were filtered out from further analysis accordingly to the lumi package standards [25]. To identify differentially expressed genes, based on a moderate t-test, the Limma package [10] was used and genes were selected based on a p-value cut-off <0.05 (after FDR adjustment) and absolute fold change >1 following the Macroarrays Quality Control Program criteria [11]. IPA software was then used (www.ingenuity.com) to detect the enrichment of biofunctions and networks in the resulting list. Core analysis was performed using the following settings: Reference set: HumanHT-12 v 4.0 [v3 is mentioned in the methods]; Relationship to include: Direct and Indirect; Includes Endogenous Chemicals; Filter Summary: Consider only molecules and/or relationships where (species = Uncategorized (e.g. chemicals) OR Human) AND (confidence = Experimentally Observed). Canonical pathways analysis has been performed using as scoring method B-H Multiple Testing Correction p-values set to a threshold value of 0.05.

Meta-analysis was performed using the CAMERA function implemented in the Bioconductor Limma package for a ‘competitive gene set test’ [12]. The resulting top Molecular signatures have been further analyzed using a ‘focused’ gene set test ROAST [27]. All the list of genes organized as Signature and present in The Molecular Signature Database C2 collection (MSigDB) (www.broadinstitute.org/gsea/msigdb/) were updated to official symbols using the illumina annotation file HumanHT-12_V3_0_R2_11283641_A, while the CR signatures were created according to the Table 2 of 3 and M.musculus to H.sapiens conversion obtained using the NCBI HomoloGene tool (http://www.ncbi.nlm.nih.gov/homologene).

Representation of differentially expressed genes in their Genomic localization and visual representation of their relationships with the different pathways were produced using CIRCOS software [28]. Statistical analysis of real time PCR and cell proliferation data was performed using Student's t test to compare the transcripts relative abundance in the CR group and WD group. P values <0.05 were considered statistically significant.