A Repeated Administration Method to Estimate Heritability of Nutrigenomic and Placebogenomic Traits

This special issue of the journal examines and highlights the new frontiers in nutrigenomics, plant omics, and food engineering. The placebo component of response to plants and nutrition, the ways in which genomic and multiomics variation contribute to placebo response, the nascent field of placebogenomics, also call for research in the near future. Yet, in emerging fields that examine the genetic basis of host responses to the environmental exposures (drugs, nutrition, or placebo), molecular genetic studies are often initiated without adequate prioritization of the highly heritable phenotypes and clinical endpoints. Hence, it is timely to examine some of the unchecked assumptions made in the field with an eye to future research and development strategy.

The extent to which phenotypic variability is heritable is a key rationale for nutrigenomics, pharmacogenomics, and placebogenomics research. Heritability studies allow for priority setting in regard to phenotypes that are best suited for genetic/genomics approaches, especially when the phenotypes of interest are complex or multifactorial, and the research funds are limited (Matthaei et al., 2015; Ozdemir et al., 2005). The missing heritability can also be inferred if, for example, a set of genetic biomarkers have been discovered and yet only partially explain the full amount of heritability associated with the phenotype of interest.

Twin studies are used to estimate heritability of person-to-person and population variability in phenotypes. Accordingly, monozygotic (MZ) and dizygotic (DZ) twin pairs are studied for concordance in rates of phenotypic prevalence. A higher concordance of the phenotype of interest in MZ over DZ twins is taken as evidence of heritability (Endrenyi et al., 1976; Vesell, 1992). But twin studies can be time consuming and costly. There are additional practical barriers to twin studies as well. Consider, for example, conducting a twin study on heritability of clinical response to drug X, used for treatment of disease Y. The study would call for recruitment of MZ and DZ twins who suffer from disease Y at the same time, and treated with drug X simultaneously, conditions that are often difficult to realize in clinical practice. Moreover, heritability of pharmacological and nutritional phenotypes differs depending on the type of drug and nutritional exposure as well as the clinical effects/endpoints in question. Dynamic traits such as nutritional, drug, and placebo responses call for new, practical, and high-throughput approaches to measure the heritability of their phenotypes.

More than two decades ago, a research team in Toronto, Canada, led by Werner Kalow and Laszlo Endrenyi, developed a high-throughput approach to estimate the heritability of drug response traits, using a repeated drug administration (RDA) method. I had the good fortune to work with both Kalow and Endrenyi as a postdoctoral fellow in Toronto in the 1990s. That was a time when there was fresh excitement (which still remains to date) to implement personalized medicine around the world, and pharmacogenomics has thus begun to transition into mainstream medical practice. There was abundance of clinically significant drug response phenotypes whose heritability remained either unknown or understudied. The RDA method involved a comparison of between- (SD_b²) and within-person (SD_w²) variances to obtain an estimate of the genetic component of variability in dynamic traits such as drug disposition. Looking through the conceptual lens of the current special issue, the RDA studies might usefully inform future plant omics, nutrigenomics, and placebogenomics research, and clinical implementation strategies by triaging the phenotypes that are most amenable to genomics/multiomics inquiries.

I shall, therefore, briefly highlight the RDA method, as described in detail elsewhere (Kalow et al., 1998, 1999; Ozdemir et al., 2000, 2005).

The independent variable in twin studies is the degree of similarity in the genetic makeup of two groups of individuals: the MZ (nearly 100% identical in the genome) and DZ twins (50% sharing of the genetic makeup on average). If a trait is highly heritable, it is reproducible to a larger extent in MZ than DZ twin pairs. The RDA method approaches heritability evaluation by comparing the variance of a dynamic dependent variable (e.g., response to a plant, foodstuff, or drug) in the same person assessed repeatedly over time (i.e., similar to the MZ twin pairs) against the variance among individuals who are unrelated to each other (Kalow et al., 1998, 1999):

“As originally proposed and demonstrated by Kalow et al. (1998), the genetic component (r_GC) of variability in a time-dependent pharmacokinetic or pharmacodynamic occurrence can be described with the following equation:

Moreover, heritability of phenotypes can display time (e.g., daytime versus nighttime) dependency as suggested in the case of the CYP3A4 drug metabolizing enzyme (Ozdemir et al., 2000). The RDA method may inform future plant omics, nutrigenomics, placebogenomics, and pharmacogenomics clinical study designs by identifying and prioritizing highly heritable phenotypes. This would enhance the likelihood of discovering robust biomarkers of individual variability in responses to plants, nutrition, drugs, and the environment.