Stitching Protein-Protien Interactions via DNA Sequencing

Stitch-Seq, one of the newest editions to the Next Generation DNA Sequencing (NGS) was presented in June's Nature Methods.

Back in 2008, when groups were realizing the power of NGS technologies, I entitled a post "Next Gen Sequencing is not Sequencing DNA" to make the point that massively parallel ultra-high throughput DNA sequencing could be used to for quantitative assays that can measure transcriptome expression, protein-DNA interactions, methylation patterns, and more. Stitch-seq can now be added to a growing list of assays that include RNA-Seq, DNAse-Seq, ChIP-Seq, or HITS-CLIP, and others.

Stitch-seq explores the interactome, a term used to describe how the molecules of a cell interact in networks to carryout life's biochemical activities. Understanding how these networks are controlled through genetics and environmental stimuli is critical in discovering biomarkers that can be used to stratify disease and target highly specific therapies. However, the interactome is complex; studying it requires that interactions can be identified at high scale.

Many interactome studies focus on proteins. Traditional approaches involve specially constructed gene reporter systems. For example, in the two-hybrid approach, a portion of a protein encoding gene is combined with a gene fragment containing a DNA binding domain of a transcription factor (bait). In another construct a different protein encoding region is combined with the RNA polymerase binding domain fragment of the same transcription factor (prey).

When the DNA constructs are expressed, interactions can be measure by gene expression. If the protein attached to the bait interacts with the protein attached to the prey, transcription is initiated at the reporter gene.  When reporter genes confer growth on selective media, interacting protein encoding segments can be identified by isolating the DNA from growing cells and sequencing the DNA constructs.

Therein lies the rub

Until now, interactome studies combined high-throughtput assays systems with low-throughput characterization systems that PCR amplified the individual constructs and characterized the DNA by Sanger sequencing. Yu and colleagues overcame this problem by devising a new strategy that put potential interacting domains on common DNA fragments, via "stitch-PCR" to prepare libraries that can easily be sequenced by NGS methods.  Using this method the team was able to increase overall assay throughput by 42% and measure 1000s of interactions.

While still low-throughput relative to the kinds of numbers were used to on NGS, increasing the throughput of protein interaction assays is an important step toward making systems biology experiments more scalable. It also adds another Seq to our growing collection of Assay-Seq methods.

Yu, H., Tardivo, L., Tam, S., Weiner, E., Gebreab, F., Fan, C., Svrzikapa, N., Hirozane-Kishikawa, T., Rietman, E., Yang, X., Sahalie, J., Salehi-Ashtiani, K., Hao, T., Cusick, M., Hill, D., Roth, F., Braun, P., & Vidal, M. (2011). Next-generation sequencing to generate interactome datasets Nature Methods, 8 (6), 478-480 DOI: 10.1038/nmeth.1597

21st Century Medicine: A Question of Ps

Last Sunday and Monday (5/15, 5/16/11) the Institute for Systems Biology (ISB) held their annual symposium. This year was the 10th annual and focused on "Systems Biology and P4 Medicine."

For those new to P4 medicine, the Ps stand for Personalized, Predictive, Preventative, and Participatory. P4 medicine is about changing our current disease oriented, reactive, approaches to those that prevent disease by increasing the predictive power of diagnostics. Because we are all different, future diagnostics need to be tailored to each individual, which also means individuals need to be more aware of their health and proactively participate in their health care. The vision of P4 medicine is that it will not only dramatically improve the quality of health care, it will significantly decrease health care costs. Hence, some folks add additional Ps to include payment and policy.

P4 medicine is an ambitious goal. In Lee Hood's closing notes he noted four significant challenges that need to be overcome to make P4 medicine a practical reality:

1. IT challenges. In addition to working on how to transform datasets containing billions of measurements into actionable information, we need to integrate high dimensional data a wide variety of measurement systems. Reduced data will need to be presented in medical records that can be easily accessed and understood by health care providers and participants.
2. Education. Students, scientists, doctors, individuals, and policy makers need to learn and develop an understanding of how the networks of interacting proteins and biochemicals that make us healthy or sick are regulated by our genomes and respond to environmental factors. 
3. Big vs Small Science. Funding agencies are concerned with how to best support the research needed to create the kinds of technologies and approaches that will unlock biology's complexity to develop future diagnostics and efficacious therapies. Large-scale projects conducted over the past 10 years have made it clear that biology is extremely complex. Deciphering this complexity requires that we integrate production-orientated data collection approaches, that develop a data infrastructure, with focused research projects, run by domain experts, that explore specific ideas.  The challenge is balancing big and small science to achieve high impact goals. 
4. Families. Understanding the genetic basis of health and disease requires the research be conducted on samples derived from families rather than randomized populations. Many families are needed to develop critical insights. However, in the U.S. our IRB (Institutional Review Boards) are considered a hinderance to enrolling individuals. 

Through the day and half conference numerous presentations explored different aspects of the above challenges. Walter Jessen at Biomarker Commons has created excellent summaries of the first and second day's presentations. For those who like raw data, the #ISB2011P4 hashtag can be used to get the symposium's tweets.