You may have heard the analogy “beads on a string” to describe genetic code. Two new papers from the HudsonAlpha Institute for Biotechnology report that both the beads and string contribute to how genetic code relates to human health.

Envision DNA as a very long string, wrapped around millions of beads made of proteins. To regulate genes, cells use thousands of different proteins.  Imagine the beads are made of thousands of combinations of different colors and designs. The technique used in these papers, chromatin immunoprecipitation or ChIP-seq, allows researchers to go in and pick out the specific protein-DNA complexes, or individual beads from this huge jumble, that they want to study.

Writing in the journal Nature Methods, Yiwen Chen and colleagues from multiple institutions, including HudsonAlpha, reported in April 2012 on the best methods to perform ChiP-seq. Since this technique is used in laboratories worldwide, the group investigated the factors most influencing accuracy of ChIP-seq, meaning they pull out the correct beads more frequently. This paper should serve as a guide for standardizing results throughout the field, allowing for easier comparisons between large datasets.

Separately, Timothy Reddy, Ph.D. and other scientists in the laboratory of Rick Myers, Ph.D., at HudsonAlpha, along with colleagues at Duke University and the California Institute of Technology, used ChIP-seq to examine the effects of protein binding on gene expression. We all should have two copies of each gene in our genome:  one from our maternal chromosome, and one from our paternal. These genes are often expressed due to the binding of specific proteins called transcription factors at the start of the gene.

Using cells from one person, the scientists were able to compare the expression of each parent’s given copy of a number of different genes in the person, and then use ChIP-seq to tell which transcription factors were bound to the genome at each copy of the gene. They saw that in 5.5 percent of genes, the levels of binding and expression were not equal between the two copies of the genes. This would mean that the individual would express more of either the paternal or maternal gene, potentially influencing many traits or risk for disease.

In particular, these sites that were differentially bound and expressed were more often associated with risk for autoimmune disease. These findings suggest an individual’s genotype, or set of genomic variants, alone, might not be enough to determine phenotype, or collection of expressed traits and disease risk.

Together, these two papers demonstrate the most accurate methods for investigating regulation of our genomic code through ChIP-seq, and the importance of understanding protein binding to DNA for proper interpretation of  DNA’s influence on human health.

The paper by Yiwen Chen et al. was published online by the journal Nature Methods on 22 April 2012, and can be found here:
http://www.nature.com/nmeth/journal/vaop/ncurrent/full/nmeth.1985.html

The paper by Timothy Reddy et al. was published in the journal Genome Research on 2 February 2012, and can be found here:
http://genome.cshlp.org/content/early/2012/03/07/gr.131201.111.abstract