Transcriptomics in Nutritional Research

The classical gene analysis approach, such as Northern blotting and real‐time RT‐PCR, can only analyze gene expression for a limited number of candidate genes at a time. DNA microarray technology allows us to measure the expression level of thousands of genes, or even entire genomes, simultaneously. A typical DNA microarray experiment includes a number of characteristic steps.

1. RNA extraction from a sample; 

2. Reverse transcription of the RNA to obtain complementary DNA (cDNA) and labeling of the cDNA with specific dyes (usually fluorophores like Cyanine 3 and 5), or reverse transcription of the cDNA to obtain cRNA and labeling of the cRNA; 

3. Hybridization of the labeled cDNA or cRNA onto the microarray under given conditions; 

4. Washing the slides to remove non‐hybridized labeled oligonucleotides; 

5. Using an appropriate scanning device to detect signal; and 

6. Data analysis by bioinformatics tools. 

There are more and more examples of DNA microarray technology being performed in cell culture systems or laboratory animals to identify the cellular responses to dietary constituents and their molecular targets. For example, green tea catechins (McLoughlin et al., 2004; Vittal et al., 2004), soy isoflavones (Herzog et al., 2004), polyunsaturated fatty acids (Kitajka et al., 2004; Lapillonne et al., 2004; Narayanan et al., 2003), vitamins D and E (Johnson and Manor 2004; Lin et  al., 2002), quercetin (Murtaza et  al., 2006), arginine (Leong et  al., 2006), anthocyanins (Tsuda et  al., 2006), and hypoallergenic wheat flour (Narasaka et al., 2006). 

For example, Lavigne et al. (2008) used a DNA oligo microarray approach to examine effects of genistein [dont confuse Genistein is a polyphenolic isoflavine that belongs to the flavonoid group and is commanly found in various dietary vegetables such as soyabeans and fava beans Genistein is claimed to exert many beneficial effects on health, such as protection against osteoporosis, reduction in the risk of cardiovascular disease, alleviation of postmenopausal symptoms and anticancer properties.] 

On global gene expression in MCF‐7 breast cancer cells. They found that genistein altered the expression of genes belonging to a wide range of pathways, including estrogen‐ and p53‐mediated pathways. At physiologic concentrations (1 or 5 μM), genistein elicited an expression pattern of increased mitogenic activity, while at pharmacologic concentrations (25 μM), genistein generated an expression pattern of increased apoptosis, decreased proliferation, and decreased total cell number. Park et al. (2008) performed a comprehensive analysis of hepatic gene expression in a rat model of an alcohol‐ induced fatty liver using the cDNA microarray. It was found that chronic ethanol consumption regulated mainly the genes related to the processes of signal transduction, transcription, immune response, and protein/amino acid metabolism. For the first time, this study revealed that five genes (including beta‐glucuronidase, UDP‐glycosyltransferase 1, UDP‐ glucose dehydrogenase, apoC‐III, and gonadotropin‐releasing hormone receptor) were regulated by chronic ethanol exposure in the rat liver.

Furthermore, the number of microarray‐based transcriptomics analysis for assessing the biological effects of dietary interventions on human nutrition and health is steadily increasing. van Erk et  al. (2006) investigated the effect of a high‐ carbohydrate (HC) or a high‐protein (HP) breakfast on the transcriptome of human blood cells with RNA samples taken from eight healthy men before and 2 h after consumption of the diets. About 317 genes for the HC breakfast and 919 genes for the HP breakfast were found to be differentially expressed. Specifically, consumption of the HC breakfast resulted in differential expression of glycogen metabolism genes, and consumption of the HP breakfast resulted in differential expression of genes involved in protein biosynthesis. Using GeneChip microarrays, Schauber et  al. (2006) examined the effect of regular consumption of the low‐digestible and prebiotic isomalt and the digestible sucrose on gene expression in rectal mucosa in a randomized double‐blind crossover trial with 19 healthy volunteers over 4 weeks of feeding. 

They revealed that dietary intervention with the low digestible isomalt compared with the digestible sucrose did not affect gene expression in the lining rectal mucosa, although gene expression of the human rectal mucosa can reliably be measured in biopsy material. Mangravite et al. (2007) used expression array analysis to identify the molecular pathways responsive to both caloric restriction and dietary composition within adipose tissue from 131 moderately overweight men. They found that more than 1000 transcripts were significantly downregulated in expression in response to acute weight loss. The results demonstrated that stearoyl‐ coenzyme A desaturase (SCD) expression in adipose tissue is independently regulated by weight loss and by carbohydrate and saturated fat intakes, and SCD and diacylglycerol transferase 2 (DGAT2) expression may be involved in dietary regulation of systemic triacylglycerol metabolism. Kallio et al. (2007) assessed the effect of two different carbohydrate modifications (a rye‐pasta diet characterized by a low postprandial insulin response and an oat‐wheat‐potato diet characterized by a high postprandial insulin response) on subcutaneous adipose tissue (SAT) gene expression in 47 people with metabolic syndrome. They detected that there are rye‐pasta diet downregulated 71 genes (linked to insulin signaling and apoptosis) and oat‐wheat‐potato diet up‐regulated 62 genes (related to stress, cytokine‐chemokine‐mediated immunity, and the interleukin pathway). 

Using microarray analysis, Niculescu et al. (2007) investigated the effects of dietary soy isoflavones on gene expression changes in lymphocytes from 30 postmenopausal women. They indicated that isoflavones had a stronger effect on some putative estrogen‐responsive genes in equol producers than in nonproducers. In general, the gene expression changes caused by isoflavone intervention are related to increased cell differentiation, increased cAMP signaling and G‐protein‐coupted protein metabolism and increased steroid hormone receptor activity. Rcently, using transcriptomics, Marlow et  al. (2013) investigated the effect of a Mediterranean‐inspired diet on inflammation in Crohn’s disease patients. They observed significant changes in gene expression, totally, 1902 genes were up‐regulated and 1649 genes were downregulated, after a 6‐week diet intervention. By Ingenuity Pathway Analysis (IPA), key canonical pathways affected by diet intervention were identified, including EIF2 signaling, B‐cell development, T‐helper cell differentiation, and thymine degradation. Rosqvist et  al. (2014) performed transcriptomics to investigate liver fat accumulation and body composition after overfeeding saturated (SFA) (palm oil) or n‐6 polyunsaturated (PUFA)(sunflower oil) for 7 weeks in 39 young and normal‐weight individuals. 

The results revealed that SFA markedly increased liver fat compared with PUFA, and PUFA caused an almost three‐fold increase in lean tissue than SFA. The differentially regulated genes were involved in regulating energy dissipation, insulin resistance, body composition, and fat cell differentiation. However, there are some problems or limitations for transcriptomics approaches in nutritional research. One major problem is non‐reproducibility of gene expression profiles. Different conclusions could be drawn from the same experiment but performed at different times or different labs or different platforms. Fortunately, for reducing errors or variations, standards for reporting microarray data have been established under MIAME (minimum information about a microarray experiment) (Brazma et al., 2001). Barnes et al. (2005) evaluated the reproducibility of microarray results using two platforms, Affymetrix GeneChips and Illumina BeadArrays. 

The results demonstrated that agreement was strongly correlated with the level of expression of a gene, and concordance was also improved when probes on the two platforms could be identified as being likely to target the same set of transcripts of a given gene. Another major issue is the analysis of the data sets and their interpretation. Analyses only providing gene lists with significant p‐values are insufficient to fully understand the underlying biological mechanisms, a single gene that is significantly upregulated or downregulated does not necessarily have any physiological meaning (Kussmann et  al., 2008). 

The combination of statistical and functional analysis is  appropriate to facilitate the identification of biologically relevant and robust gene signatures, even across different microarray platforms (Bosotti et al., 2007). An additional and more specific limitation in human nutritional applications is that microarray studies require significant quantities of tissues material for isolation of the needed RNA, while access to human tissues is obviously limited, although it is not impossible to obtain biopsies from a control subjects involved in a nutrition research. If using human blood cells instead of tissue material, large inter‐individual variation exists in gene expression profiles of healthy individuals (Cobb et al., 2005), this makes it challenging to identify robust gene expression signatures in response to a nutrition intervention. On the other hand, sample handling and prolonged transportation significantly influences gene expression profiles (Debey et al., 2004), the highly standardized protocol across different labs is needed. In particular whole‐blood samples require the depletion of globin mRNA for enabling detection of low‐abundance transcripts. Shin et al. (2014) showed that the experimental globin depletion removed approximately 80% of globin transcripts, and allowed for reliable detection of thousands of additional transcripts. However, a concern is that globin depletion leads to the significant reduction in RNA yields.