Jim Kaput talks with
ScienceWatch.com and answers a few questions about
this month's Fast Moving Front in the field of Biology
& Biochemistry. The author has also sent along
images of their work.
Article: The case for strategic international
alliances to harness nutritional genomics for public and
personal health
Authors: Kaput,
J, et al.
Journal: BRIT J NUTR, 94 (5): 623-632 NOV 2005
Addresses: Univ Calif Davis, Ctr Excellence Nutr Genom,
Davis, CA 95616 USA.
Univ Calif Davis, Ctr Excellence Nutr Genom, Davis, CA
95616 USA.
Tufts Univ, USDA, Human Nutr Res Ctr Aging, Nutr &
Genom Lab, Boston, MA 02111 USA.
(addresses have been truncated)
Why do you think your paper is highly
cited?
More and more scientists are realizing that an understanding of complex
biological systems cannot be achieved by analyzing merely one aspect of a
particular problem. Nutritionists typically ignore genetic variation, and
many genetic association studies do not measure environmental influences.
This paper was a call to scientists of many disciplines to collaborate
under the banner of nutrigenomics—although the more popular term used
now is personalized nutrition.
The genesis of this paper was initiated at the 1st Bruce Ames Symposium on
Nutritional Genomics that was held at the University of California, Davis,
in October, 2004. A group of about 20 scientists at that multidisciplinary
meeting met and agreed that understanding gene-nutrient (or
gene-environment) interactions would require the knowledge and technologies
of many disciplines and international alliances. This group formed the core
that developed the concepts in the paper. These concepts were rooted in the
studies in humans and laboratory animals conducted in the 1990s and early
2000s which demonstrated the need for analyzing multiple genotypes and
multiple diets in order to understand biological processes.
Since genotypes and cultures vary across the world, it was not a great leap
to realize that the study of gene-nutrient-phenotype associations could
only be done through international collaborations. To accomplish such an
ambitious goal, we outlined the need for data sharing, improving the
analyses and consistency of phenotypes (deep phenotyping), the need for
genomic controls (population architecture analyses), ethical issues, other
environmental variables, and more accurate nutritional assessments.
While some of these needs were the same ones for the International
Haplotype Mapping project, phenotyping and assessing nutritional intakes
added to the complexity of the task. Our colleagues in nutrigenomics
organizations in Europe: European Nutrigenomics Organization,
New Zealand,
Brazil, Canada, and other countries are all involved
in meeting these challenges.
In addition, the human variome project is now collaborating with us
on some aspects of our effort and we on their effort. Progress has been
made on many of these topics, and many are focusing on various needs
described in the paper, citing our publication in the process.
Does it describe a new discovery, methodology, or
synthesis of knowledge?
The paper highlighted the re-awakening of long-known knowledge: that what
an individual eats will affect one's health (Hippocrates: "Let medicine be
your food and food your medicine.") The only difference from that ancient,
but largely ignored, knowledge is the ability in the modern era to conduct
high-throughput analyses of biological systems using various omic
technologies. Many of us acknowledge that nutrigenomics sounds as if the
only omic is genomic, when in reality, analyzing transcripts, proteins, and
metabolites are required for a complete understanding of the effects of
nutrients (and lifestyle) on an individual's genetic make-up.
This paper was actually one of many reviews of the early 2000s which began
summarizing the need to study nutrient intakes and genetic variation. In
one sense, its appeal and utility was needed to synthesize the knowledge of
these scientific fields, along with the fact that 89 authors from 22
countries agreed to what was needed to make progress. Many of these authors
are still working together, both formally and informally, to forge new
initiatives in this area of research. It goes without saying that many of
these authors could have produced the first draft of this article and that
many contributed excellent edits which yielded the final version.
Would you summarize the significance of your
paper in layman's terms?
The general public knows that food is important for health, but many are
confused by the nutritional epidemiological studies that yield conflicting
conclusions about what to eat. The general hype surrounding genetic
research has led to a deterministic view of disease susceptibility: the "It
does not matter what I eat or how active I am, my genes dictate my health
outcomes" syndrome.
Our paper acknowledged the limitations in current research methods and laid
out a specific plan to improve the science. Our goal is to understand why
an individual such as Jim Fixx died at age 52, even though he was a runner
and ate a healthy diet, yet the English statesman Winston Churchill lived
to age 90 while being overweight and smoking cigars throughout his life.
How did you become involved in this research
and were any particular problems encountered along the way?
A well-known and highly respected nutritionist, Dr. Willard Visek, M.D.,
Ph.D., of the University of Illinois, Urbana-Champaign, passed by my office
each day, often entreating me to teach him about molecular biology. In his
view, the discipline of nutrition needed more concrete answers to questions
of how nutrients influence health. Although I was reluctant to change
fields as a young assistant professor in the Biochemistry Department, I
realized that the emerging tools of high-throughput molecular biology could
be used to analyze nutrient–gene interactions.
Although the molecular methods we started using on this project in 1986
were crude by today's standards, emerging technologies and ideas were
transforming how biological systems were being analyzed. Some rather brave
graduate students (Tim Elliot, Deborah Swartz, Liz Paisely, Eric Park) and
one of the best research associates, Heather Mangian, started a series of
studies that helped developed a mouse model for studying gene-nutrient
interactions.
"The list of potential implications
and applications of this science are quite
lengthy and extensive."
Our strategy for identifying these gene-nutrient interactions was to
compare gene expression—which could be at the metabolite or protein
levels—among two or more strains of mice that (i) differ in response
to the same diets and (ii) had different genetic susceptibility to disease.
Only the differences in gene expression could produce a different phenotype
(disease or not) based on diet.
We proposed (see Kaput J, et al., "Diet-disease interactions at
the molecular level: An experimental paradigm," J. Nutr. 124:
1296S-1305S, 1994) that genes regulated by diet and mapping to genetic loci
involved in disease would identify the likely causal genes, i.e., within
quantitative trait loci, (see Park EI, et al., "Lipid level and
type alter stearoyl CoA desaturase mRNA abundance differently in mice with
distinct susceptibilities to diet-influenced diseases," J. Nutr.
127: 566-73, 1997, and Kaput J, et al., "Identification of genes
contributing to the obese yellow Avy phenotype: caloric restriction,
genotype, diet x genotype interactions," Physiological genomics
18: 316-24, 2004). Subsequent studies by others called these "expression
quantitative trait loci," or eQTLs.
Our work unknowingly was consistent with the research of pioneers such as
Charles Scriver, the eminent Canadian pediatrician and biochemical
geneticist, and others, who demonstrated that the avoidance of
phenylalanine prolonged life and also reduced symptoms for those with
phenylketonuria—that gene-environment interactions contribute to
health and disease processes.
Others were showing that different fat intakes in humans produced different
serum lipid levels in individuals with polymorphisms in apolipoproteins
(see Corella and Ordovas, "Single nucleotide polymorphisms that influence
lipid metabolism: Interaction with dietary factors," Annu. Rev.
Nut., 25:341-90, 2005). Even with the work of many in humans and
laboratory animals, nutrigenomic ideas , concepts, and experiments (and our
grant applications) were not well accepted in the 1980s and 1990s.
Where do you see your research leading in the future?
The goal of the authors on this paper, and our new colleagues and
collaborators around the world, is to produce more complete answers about
how nutrients influence health in the individual. That latter statement is
still not well understood by all scientists or the general public.
Our current experimental designs for nutritional or genetic epidemiology
are based on population studies. These studies yield the population's
attributive risk (see Kaput J, "Nutrigenomics research for personalized
nutrition and medicine," Current Opinions in Biotechnology 19:
110-20, 2008) and not individual risk factors.
While the population-attributable risk (PAR) provides useful guidelines for
what to eat or which genes and their variants may contribute to health or
disease, epistatic interactions and other gene-environment interactions may
alter the effect of a single nucleotide polymorphism (SNP) in different
individuals. One approach that is based on the concept of challenging
homeostatic systems (e.g., the oral glucose tolerance test, measuring
glucose but also other serum metabolites—see van Ommen B, et
al. "Challenging homeostasis to define biomarkers for nutrition
related health," Molecular Nutrition and Food Research, in press).
While nutriomic research provides the technologies and concepts, we need to
develop novel research strategies for developing individual risk factors,
while taking into account gene variants, epistatic interactions resulting
from differing genetic ancestries, and influences of different
environments—quite a challenging task.
Do you foresee any social or political implications for your
research?
The results of nutrigenomic research have significant implications for
society, from food fortification programs that account for the majority
genotype in a region (see Darnton-Hill et al., "Public health
nutrition and genetics: implications for nutrition policy and promotion,"
Proc. Nutr. Soc. 63: 173-85, 2004), to how drugs are developed and
used. For example, agribusinesses interested in developing plants with
improved nutrient profiles will need to know which populations will benefit
from their products and which populations may be harmed.
Some dietary chemicals which alter drug metabolism pathways and basic
dietary profiles are rarely considered in drug treatments, although
specific drug-diet interactions are well known. Food manufacturers may
adopt research results to create healthy, stable, and economical foods
based on health-related evidence rather than relying only on taste and
shelf life. The list of potential implications and applications of this
science are quite lengthy and extensive.
Jim Kaput, Ph.D.
Director
Division of Personalized Nutrition and Medicine
FDA/National Center for Toxicological Research
Jefferson, AR, USA
Web
Many nutritional or genetic epidemiology studies analyze a population of
(largely) one ancestral group (e.g., Europeans - such as the group
represented by the column on the right in the top graph) and apply those
results to the rest of the human population (the bell shaped curve). Only a
small sampling of the total genetic variation, phenotypic variation, or
lifestyle (diet) variation is accounted for in such studies. International
collaborations are needed to find populations with the greatest differences
in gene - environment interactions. That is, instead of examining the
statistical main effect of genetic variation differences between
populations or the statistical main effect of different nutrient intakes,
the statistical main effect that is most important is the interaction term
(gene - environment). The first step, therefore, is to conduct a harmonized
protocol so that gene - environment interactions can be analyzed to find
those of greatest difference. While one may predict these to correspond to
difference between different ancestral groups, variables such as gene -
gene interactions may confound simplistic experimental designs. Once these
extremes are found, the intermediate bins can be filled in using standard
statistical concepts (quintiles, quartiles etc). The goal of this design is
to discover the full range of human gene - environment interactions by
finding and then defining metabolic groups that respond most differently.
(see Current Opinions in Biotechnology 19: 110 – 20, 2008)
for more details.
KEYWORDS: STRATEGIC INTERNATIONAL ALLIANCES; NUTRIGENOMICS;
GENE-NUTRIENT INTERACTIONS; HEALTH DISPARITIES.