We are what we eat. The connections between our diets and the risk of chronic disease have become increasingly clear over the past 30 years of research. This has led to many recommendations from the National Institutes of Health - these include the five-a-day program to increase the amounts of fruits and vegetables in our diet, as well as pronouncements regarding reducing the fat content of the diet.
The risks for cancer. Although there are certain mutation in genes that substantially increase the risk of cancer (e.g., BRCA1 and p53), only 5% of cancers are attributable to these heritable genes. The remaining risk comes from factors such as exposure to cancer causing agents (e.g., polycyclic hydrocarbons from environmental exposure and industrial chemicals), from lifestyle (e.g., diet, obesity, lack of physical exercise, some medical treatments), or physical characteristics (e.g., internal hormone levels).
Can diet prevent cancer? Foods contain macro components that give us the calories for our energy and the amino acids for our muscles and micro components (i.e., vitamins and minerals) that provide the nutrition necessary to maintain health. Foods also contain many other minor but very important compounds that can't be made by the body and haven't been well studied. These compounds include the bioflavonoids. International epidemiological studies have found that certain edible plants (fruit/vegetables) are an important part of the diet in areas where there are low risks of individual cancers. This has led to the extraction of individual compounds from the plants to evaluate their biological activities in cancer prevention.
Bioflavonoid-containing foods. Many foods contain bioflavonoids - those that have been shown to have some measure of cancer prevention properties include isoflavonoids from soy, resveratrol from grapes and wine, and catechins in green tea.
Chemoprevention models. Rodent models of breast and prostate cancer have been heavily used at UAB by Clinton Grubbs, Coral Lamartiniere and Stephen Barnes to test the chemopreventive properties of chemical and nutritional agents. In Project 1, Coral Lamartiniere will examine the effects of three polyphenols (genistein, resveratrol and green tea polyphenols) singly and in combination on the incidence of mammary tumors in animals treated with the carcinogen 7,12-dimethylbenz[a]anthracene at 50 days of age. The polyphenols will be administered orally in the prepubertal period. Mammary tissue will be obtained during puberty and at the time of sacrifice in order to carry out DNA microarray (Core B) and proteomics analysis (Core C).
Genes that interact with nutrients. Foods provide the components that are needed to create DNA, RNA and proteins and the enzymes that catalyze their formation. However, they can also act directly on individual parts of the biosynthetic systems, particularly to alter the function and activities of transcription factors, or to increase/decrease methylation of critical regions of the genome.
Role of gene variants in cancer. Population analysis of the sequences of individual genes has revealed that there are many single nucleotide sites where a mutation has been stably incorporated. These single nucleotide polymorphisms do not cause catastrophic changes that lead to genetic diseases in the young. Rather than causing the loss of the activity of the protein or its severe truncation, the protein may only have reduced activity or lack an intracellular targeting sequence. Some of these gene variants may be associated with an increased risk of cancer. The concept that these SNP mutations are "not inheriently good or bad but just different between people" like blue eyes and brown eyes.
Diet, gene variants, physical development, and cancer risk. Since dietary components, such as the polyphenols, can alter the synthesis of genes and subsequent gene expression, there is the possibility that choosing the right diet to go with a given set of genes may help to prevent the expression of genes that increase the risk for cancer or enhance the expression of genes that reduce cancer risk. As a measure of this, Pam Horn-Ross in Project 2, the GRowth and LifeStyle (GRLS) Study, will examine whether girls, age 10-13, who consume soy foods in amounts equivalent to a typical Asian American diet develop at a different rate than girls who consume a typical western-style diet.
High dimensionality of CNGI data. Analysis of traditional laboratory experiments has typically involved monitoring 1-5 different variables. The statistical analysis of such systems is well-described and is available through programs such as SAS. However, nowadays the type of research carried out in CNGI-sponsored experiments (see genes and cancer risk and proteins and cancer risk) contain 100 to 1,000 to 10,000 and higher variables. This presents a considerable challenge for statistical analysis. Put simply, there is no closed form for the estimation of the statistical probabilities of each of the variable factors. David Allison's group in Project 3 will focus their efforts on development of novel procedures and accompanying software to carry out statistical analysis of high-dimensional data.
Genes and cancer risk. The sequencing of the human genome and the genomes of animals used as models of cancer has led to the generation of microarrays that contain 10,000 or more different genes. These can be used to interrogate the transcriptosome, the family of mRNAs produced by DNA transcription. This is carried out by converting the mRNA to cRNA and attaching different fluorescent dyes to the control and treated samples. The ratio of the fluorescent dyes that bind to each DNA spot provides a measure of the relative expression of individual genes. Analysis of DNA microarrays and single nucleotide polymorphisms in this project will be carried out in Core B.
Proteins and cancer risk. Emerging technologies are now allowing investigation of the proteome - the family of proteins produced in a given cell or tissue in the cancer model. These are based on the separation of proteins by high-resolution two-dimensional electrophoresis and by two-dimensional liquid chromatography of proteolytic fragments of the proteome or proteome fraction. In both cases, the proteins are analyzed using mass spectrometry - either matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF), or nanoLC-electrospray ionization tandem mass spectrometry on a quadrupole-orthogonal time-of-flight instrument, or a Fourier transform-ion cyclotron resonance instrument. Proteomics analysis will be carried out in Core C.
Biostatistical analysis. Analysis of data arising from CNGI-sponsored projects will be carried out using programs accepted for use in high-dimensional research. New approaches and the associated software generated in project 3 will be made available to the Biostatistics Core group (Core D) and to users in other NCI-sponsored centers.
Bioinformatics. A crucial part of an endeavor of this size (in association with other NCI-sponsored centers) is to make all the data available to investigators in the broader scientific community. It is appreciated that finding solutions in the data that will be collected by CNGI investigators may strongly depend on the energies and expertise of investigators of our center. Core D will endeavor to present data from CNGI experiments in web-accessible form within 6-months of its original collection.