Tocotrienol-Rich Fraction Modulates Genes Expression in Oxidative Stress-induced Caenorhabditis e legans

Objective: The study was done to determine the effect of Tocotrienol rich fraction (TRF) on the expression of RNAs in C. elegans under oxidative stress. Methods: The nematodes were divided into 4 groups and treated accordingly: control; TRF; hydrogen peroxide (H202); TRF treatment before and after H202-induction (TRF+H202+TRF). Expressions of RNAs were analyzed with Affymetrix Genechip C. elegans Genome Array and Genespring GX11 software where differentially expressed genes were further analyzed using gene ontology (GO). Selected genes (unc-15, cit-1.2, ftn-1, rsks-1, unc-4 and daf-12) were analyzed with RT-qPCR to validate the results. Results: TRF modulated the expression of 314 genes involved in determination of adult lifespan, regulation of growth and lipid modification. A total of 440 genes involved in RNA metabolic processes, transcription, growth and differentiation of muscle and nerve cells were differently expressed following H202 induction. TRF treatment before and after H202induction resulted in 438 differentially expressed genes involved in RNA metabolic processes, transcription, response to xenobiotic stimulus and protein amino acid phosphorylation. Conclusion: TRF modulates the expression of genes involved in the regulation of lifespan in C. elegans.


Introduction
One unique feature in organism life cycle is the aging process. It occurs when the physiological system experiences a progressive functional decline which leads to poor maintenance of homeostasis and consequently death 1 . Many studies have been conducted to investigate the process and a number of theories have emerged to explain the underlying mechanism. Of the many theories, the free radical theory of aging is widely accepted because many experimental studies conducted since its introduction supported the hypothesis that free radicals cause senescence. Hydrogen peroxide (H 2 O 2 ) is one of the most abundant ROS in living cells 2 . It is produced as a by-product of aerobic metabolism and it has been demonstrated to be involved in apoptosis pathway, induction of intracellular oxidative stress and acceleration of aging 3,4. It has been shown in many organisms ranging from invertebrates to humans that the increase in oxidative damage to lipid, protein and DNA correlated with increasing age 5,6 . Among major model organism, nematode Caenorhabditis elegans (C. elegans) is well-suited for the aging research. C. elegans offers great advantage as a well-established model due to its short life cycle, large progeny production in a short time, ease of maintenance in laboratory as well as morphological simplicity which facilitates the aging study 7 .Moreover, the completely sequenced genome of this nematode makes it preferable for the molecular genetic analysis to be carried out 8 .There is also a strong conservation between C. elegans and mammals in cellular and molecular principles where 60-80% of human genes have been identified in C. elegans 9 . Many genes have been identified to control aging in C. elegans 10 . For example, mutants in age-1 gene which encodes class-I phosphatidylinositol 3-kinase (PI3K), have been found to have long lifespan, increased resistance to stress and impaired reproduction 11 . Elsewhere, nematodes with mutation in the insulin/IGF/daf-2 signaling pathway had prolonged lifespan as compared to the wild type 12 . Under normal conditions, daf-2 gene in C. elegans has also been identified to signal the AGE-1 PI3K, PDK-1 and AKT-1/2 kinases to negatively regulate daf-16, a forkhead transcription factor FOXO which is essential for longevity and stress resistance. The function of this pathway in mediating longevity and metabolism is also well-conserved in Drosophila sp. and mammals 13,14 . Over the years, there has been a growing effort among researchers to understand the genetic basis of aging which shows promises that lifespan and aging process can be genetically manipulated. The search for compounds that can delay aging and extend lifespan in model organism is widely conducted. Therefore, studies on modulation of endogenous antioxidant defences with antioxidants supplementation can be regarded as promising strategies to delay aging. Vitamin E is a lipid-soluble antioxidant which can be found naturally in vegetable oil and lipid-rich plant product 15 .Vitamin E occurs in nature in eight different isoforms: α-, β-, γ-and δ-tocopherols and α-, β-, γ-and δ-tocotrienols. Tocotrienols differ from tocopherols in that they have an isoprenoid rather than a saturated phytyl side chain 16 . It is well-accepted that tocotrienols is a more potent antioxidant as compared to tocopherols 17,18 . Apart from its antioxidant properties, tocotrienols are beneficial to health due to its neuroprotective effect, anti-carcinogenic as well as cholesterollowering property [19][20][21] . It has also been demonstrated that tocotrienols play a role in regulating signal transduction of cell death pathway 22 . However, the molecular mechanism of how tocotrienols modulate the aging pathway is still scarce. Adashi and Ishii (2000) reported that tocotrienols reduced the accumulation of protein carbonyl and consequently extended the mean lifespan of C. elegans 23 . In our previous study, TRF at a concentration of 50 µg/ml was found to restore the mean lifespan and reduce the accumulation of lipofuscin in oxidative stressinduced C. elegans 24 . Based on these results, it was hypothesized that TRF treatment stimulated cellular regeneration that lead to restoration of lifespan in C. elegans by modulating the expression of genes in various metabolic pathways. To elucidate the molecular mechanism of lifespan extension by this antioxidant, the microarray technology is further used to analyze the changes in gene expression with TRF treatment in the present study.

Nematode Strain and Culture Conditions
The wild type C. elegans strain (N2) were grown at 20°C on nematode growth medium (NGM), with E. coli OP50 as food source according to the method by Brenner (1974) 26 . All of the maintenance and handling procedures for the nematodes were conducted as described previously 26 .

H 2 O 2 -induction of oxidative stress and TRF treatment
The TRF Tri-E 70 was supplied by Sime Darby Bioganic, previously known as Golden Hope Bioganic (Selangor, Malaysia), which contains 70% of total vitamin E (15% α-tocopherol, 23% α-tocotrienol, 2% β-tocotrienol, 20% γ-tocotrienol and 11% d-tocotrienol). The optimum dose of H 2 O 2 (0.3 mM) for induction of oxidative stress, and the optimum concentration of TRF (50 µg/ml) to be treated in C. elegans were predetermined in our previous study 24 . The nematodes were divided into four groups and treated accordingly: control, H 2 O 2 induction (H 2 O 2 ), TRF treatment (TRF), and TRF treatment pre-and post-H 2 O 2 induction (TRF+H 2 O 2 +TRF). The TRF treatments were given from hatching until day 3 of adulthood to determine the ability of C. elegans to recover from H 2 O 2 -induced oxidative stress during the developmental phase.

RNA Extraction and Quantification
At day 3 of adulthood, worms were isolated from NGM agar plates and washed several times with M9 buffer (3 g KH 2 PO 4 , 6 g Na 2 HPO 4 , 5 g NaCl, 1 ml 1 M MgSO 4 , dH 2 O to 1 L) to remove any contaminating E. coli. Total RNA from each group was extracted using TRIzol reagent following manufacturer's instruction (Invitrogen, USA), purified using RNeasy Mini Kit (Qiagen, USA) and assessed for RNA quality and quantity by using NanoDrop 1000 Spectrophotometer and Agilent 2100 Bioanalyzer. Through Agilent 2100 Bioanalyzer, the RNA integrity was assessed by RNA Integrity Number (RIN) that indicates the degree of degradation of isolated RNA (RIN 1 -10; with 1 being the most degraded and 10 being the most intact) 27 . Samples with RIN of >7 were chosen for further experiments to ensure experimental accuracy.

C. elegans Whole Genome DNA Microarray
Purified total RNA was amplified and labelled with biotinylated nucleotide analog using Affymetrix GeneChip 3' IVT Express Kit according to the manufacturer's protocol. The biotin-labeled amplified RNA (aRNA) was then fragmented before being hybridized onto the genome array. The hybridization controls was used to monitor the hybridization quality, which composed of a mixture of biotinylated and fragmented cRNA of bioB, bioC, bioD and cre prepared in staggered concentrations (1.5, 5, 25, and 100pM respectively).Using Affymetrix Genechip C. elegans Genome Array, 22625 transcripts were screened for changes in expression patterns and levels. After 17 hours of hybridization, washing and scanning of the arrays was carried out using Gene Chip Hybridization, Wash and Stain Kit (Affymetrix). The performance of every chip was evaluated for indication of successful hybridization, washing and staining procedures using Affymetrix Expression Console software.

Analysis of Microarray Data
The probe intensity values generated by the Affymetrix scanner were stored in .CEL files format, and Robust Multichip Average (RMA) algorithm was applied to normalize all the data before further downstream data analyses being carried out in Genespring GX11 software. Briefly, the raw signals were log transformed and normalized using the Percentile shift normalization method, the value was set at 70th percentile. Probes with intensity values below 20th percentile were filtered out using the "Filter Probesets by Expression" option in Genespring GX11. The unpaired T-test was applied to find the candidates for differential expression, and genes with significant signal level between two different conditions (p-value cut off at 0.05) were collected. Expression ratios between pairs of compared conditions were log2 transformed. The differentially expressed genes were further subjected to hierarchical clustering analysis (Distance metric: Pearson centered; Linkage rule: Centroid) to group the samples according to the degree of similarity of their expression profile. Principle Component Analysis (PCA) was also performed to check whether samples from similar experimental condition were clustered together or not. Genes with log2 (fold change) of ≥1.2 (up-regulated) or ≤-1.2 (down-regulated) were selected as candidate genes for further functional analysis using Database for Annotation, Visualization and Integrated Discovery (DAVID) version 2.1. DAVID 2.1 tool has been widely used to elucidate the biological meanings of gene lists derived from microarray study through identification of enriched GO terms 28 . The differentially expressed genes were mapped to the Search Tool for the Retrieval of Interacting Genes (STRING) database to investigate the functional interaction among genes and proteins, thus identify the target genes for validation according to their confidence, evidence, actions or type of interaction in the protein networks.

RT-qPCR Analysis
Validation of selected differentially expressed genes (Table 4) primarily derived from microarray analysis were performed through RT-qPCR analysis (iCycler iQ5™ RT-PCR detection system, Bio-Rad, USA). Gene-specific primers were designed using the online tools, Primer3 (http://frodo.wi.mit.edu.primer3/) based on the complete mRNA sequences obtained from NCBI (http://www.ncbi.nlm.nih.gov/). RT-PCR analysis was carried out by using iQ™ SYBR Green Supermix kit (Bio-Rad, USA) and cDNA sample for each selected genes were synthesized to be used as a target for amplification step. The optimal annealing temperature for each target genes was assessed via temperature gradient feature in the iQ5™ system. The relative expression values (REV) for all target genes were calculated based on the C T value obtained from the amplification curve. Ethical clearance: This research was approved by the ethics committee of UKM Molecular Biology Institute, Kuala Lumpur, Malaysia.

Gene Expression Changes in C. elegans
The gene expression data was collected from four to five independent experiments (4-5 biological replicates), and box-whisker plot was used to visualize the data after it was normalized and summarized through RMA algorithm in Genespring GX 11 Software (Fig. 1). Samples quality was assessed through hybridization controls which in turn depicts the hybridization quality. All samples were in good quality and the hybridization process was done properly as shown by the consistency of profile intensity plots generated and there was no deviation of signal from the expected intensity profile. Comparisons of genes expression level were done between the following three sets of interpretations: Based on statistical analysis (unpaired T-test; p<0.05; log2 fold change cut-off at 1.2), we found a total of 440 genes to be differentially expressed upon induction with H 2 0 2 in which 248 genes were up-regulated by 1.2 fold and 192 genes were down-regulated by 1.2. A comparison between control and TRF group showed 314 genes to be differentially expressed where 115 genes were found to be up-regulated and 199 genes to be down-regulated. In addition, TRF treatment before and after H 2 O 2 -induction (TRF+H 2 0 2 +TRF group) resulted in 438 differentially expressed genes where 177 genes were up-regulated and 261 genes were down-regulated. Clustering analyses on differentially expressed genes showed that within group samples were clustered according to the degree of similarity of their expression profiles as shown in the dendrograms and that samples representing the same experimental condition were grouped closer together in the PCA plots (Fig. 2).

Gene Ontology Analysis on Differentially Expressed Genes through DAVID online tools
Lists of up-and down-regulated genes were uploaded to DAVID 2.1 tool. GO analysis was performed using Affymetrix probe set identifiers to calculate statistically enriched GO biological process annotations for the differentially expressed genes. The output of functional annotation analysis using DAVID 2.1 online tools appeared as a list of GO terms (mainly biological processes) that were significantly enriched (P-value <0.05). Besides that, the enrichment scores shown in the generated output represent the overall importance (enrichment) of gene groups.
Functional annotation analysis of 440 differentially expressed genes between control and H 2 O 2 group showed that 24 GO terms were significantly enriched (p<0.05) ( Table 1). Of these, the regulation of RNA metabolic process and regulation of RNA transcription involving a large number of genes (30-34 genes) were significantly affected. Interestingly, DAVID functional analysis on 314 significant genes that were changed upon TRF treatment showed that a large fraction of genes were involved in the regulation of growth and growth rate. On top of that, biological processes associated with lifespan determination, lipid modification and glycosylation, and also aging were enriched with TRF treatment ( Table 2). TRF treatment before and after H 2 O 2 induction also remarkably affected the biological processes in C. elegans. Functional analysis of 438 differentially expressed genes between H 2 O 2 and TRF+H 2 O 2 +TRF groups revealed 11 statistically significant biological processes in which most of the genes were involved in the regulation of RNA metabolic process and transcription (Table 3). Differentially expressed genes in selected biological terms were mapped to network of interactions using online tool, STRING (string-db.org) to explore the pattern of regulation of biological process based on the functional interactions between genes. Gene of interest (GOI) was selected from the interaction networks (see Supplementary data) for further validation by RT-qPCR assay ( Table 4). The finding from RT-qPCR assay indicated that all GOIs showed similar expression pattern to that of those derived from microarray experiment and thus validated the microarray data (Fig. 3).

Discussions
Nematodes react to oxidative stress condition by launching the response mechanism through the induction of gene expression. In our previous study, exposure to exogenous H 2 O 2 shortened the lifespan of C. elegans and increased the accumulation of oxidative biomarkers such as lipofuscin and 8-hydroxydeoxyguanosine 24 . Although little is known about H 2 O 2 -responsive genes in C. elegans, we believe that exposure to exogenous H 2 O 2 might regulate the changes in C. elegans genes expression based on several lines of evidences that explained its important roles in the activation of signaling pathways to stimulate cell proliferation 29,30 and differentiation in multicellular organisms 31,32 . Differential genes expression analysis using unpaired t-test in Genespring revealed that H 2 O 2 affected the transcription process of genes such as cit-1.2 which has been reported to affect the lifespan of C. elegans through RNA interference studies 33. cit-1.2 gene encodes the protein cyclin-T2 that belongs to the cyclin family 34 . It stimulates RNA and protein binding as well as regulates the enzyme activity and transcription. Cyclin T/CDK9 complex forms the transcription factor, P-TEFb that stimulates transcriptional elongation 35 . According to Kohoutek et al. (2009), a decreased in cyclin-T2 expression down-regulates the expression of genes involved in transcription and metabolism 36 . Reduced expression of cyclin-T2 was found to down-regulate a group of genes that are involved in response to stimuli and stress, muscle development, negative regulation of transcription and metabolism and signal transduction 37 . This gene has been found to have a pro-longevity effect on C. elegans 33 . Hundreds of genes have been identified to be involved in the longevity of C. elegans 38 . Though it was uncertain if tocotrienol could modulate the expression of these genes, previous studies on the molecular aspect of tocotrienol reported that it modulated genes expression in some disease models [39][40][41] . GO analysis in this study indicated that TRF affected the expression of genes involved in important age-related biological processes such as determination of adult lifespan and regulation of growth. Interestingly, many of these genes such as daf-2, dod-20, dod-21, lbp-7, rsks-1 and ftn-1 were found to be associated with aging and/or longevity in C. elegans according to the AnAge database which denotes the influence of genes to longevity based on longevity records and life history traits of organism and species 42 . TRF treatment significantly increased ftn-1 expression and decreased rsks-1 expression (p<0.05) as compared to control group.ftn-1 encodes FTN-1 or feritin which is essential for normal lifespan under iron stress conditions and has been reported to be essential for embryogenesis 34 . Previously,ftn-1 gene expression was found to be directly proportional to the lifespan of C. elegans. Mutations in ftn-1 gene was found to shorten the lifespan of C. elegans 43 , while increased expression of ftn-1 extended the lifespan of yeast 44,45 . rsks-1 encodes a putative ribosomal protein S6 kinase (S6K) that is orthologous to human p70S6K protein. rsks-1 is an age-related gene that has anti-longevity effect in which increased expression leads to shortened lifespan of C. elegans 38 . In this study, ftn-1 and rsks-1 genes were thought to affect C. elegans lifespan via several signals pathways that is triggered by the reduction of DAF-2 signaling. Expression of ftn-1 gene is coordinately regulated by insulin/IGF-1 and HIF signaling pathways, previously known to interact in the regulation of stress resistance and lifespan 46 . On the other hand, rsks-1 expression is regulated by TOR signaling which acts as a downstream target of insulin/IGF signaling that is known to be important in modulating aging and other age-related diseases 47 . The fact that TRF modulated the expression of these aging related genes known to be involved in different biological processes indicates a promising role of TRF in lifespan extention. Pre and post treatment of TRF modulated the expression of 438 genes in H 2 O 2 induced nematodes. Some of these genes have been reported to be associated with aging process in C. elegans. For instance, RNA interference of cit-1.2 gene 33 and mutation in daf-12, unc-4 and unc-26 genes were found to affect the lifespan of C. elegans [48][49][50] . Based on network interaction analysis through STRING online tool, cit-1.2 and daf-12 genes were found to be involved in the regulation of aging process. The increased expression of cit-1.2 gene with TRF treatment in H 2 O 2 -induced nematodes supports our previous finding that TRF restores the lifespan of oxidative stress-induced C. elegans 24 . In contrast with cit-1.2 gene, daf-12 gene was downregulated in TRF+ H 2 O 2 +TRF group as compared to H 2 O 2 group. daf-12 encodes a member of the steroid hormone receptor superfamily that affects dauer formation downstream of the TGF-and insulin signaling pathways 34 daf-12 is essential for normal development of nematodes, as well as responsible to ensure survival under harsh conditions and most importantly it is involved in modulating aging process in nematodes 51 . This gene showed antilongevity effect in which its activation leads to a shorter lifespan of hermaphrodites 52 . Mutations of this gene can extend the lifespan of C. elegans by almost 4 times more than normal lifespan 48 .

Conclusion
Besides having antioxidant properties that can protect against oxidative damage, TRF also modulated the expression of genes in the insulin/IGF-1 signaling pathway and its downstream pathways involved in the regulation of lifespan in C. elegans.   11 unc-26, unc-15, unc-60, hil-2, unc-89, hil-3, tbb-4,  The count indicates the number of observations from the input of the 440 genes. Terms are listed in decreasing order of significance (P-value). The count indicates the number of observations from the input of the 314 genes. Terms are listed in decreasing order of significance (P-value).