YouTube Facebook LinkedIn Google+ Twitter Xingrss  

Protein Analysis Powers Up

By Robert M. Frederickson

April 15, 2005 | While genomic and genetic systems can determine the proteomic blueprint for the machinery of the cell, it is the individual proteins themselves and their myriad interactions in various functional complexes that comprise that machinery. This cellular machinery is responsible for the replication, transcription, and translation of nucleic acid, and even the synthesis of many of its constituent components.

The invention of the two-hybrid assay for protein interaction by Stan Fields and colleagues in the early 1990s was a milestone in the study of protein interaction. This simple genetic assay in yeast was accessible to almost any laboratory that was proficient in basic yeast biology techniques. Adapted by Clontech laboratories (now part of BD Biosciences) into a commercial kit called Matchmaker, the assay has led to the rapid identification of thousands of new protein interactions within a decade.
TIES THAT BIND: Biocore T100 system determines protein binding affinities disassociation rates in real time
Over time, the two-hybrid assay was adapted to high-throughput array- and library-based mating formats, facilitated by the use of robotic instruments such as the Biomek 2000 Laboratory Automation Workstation or Tecan’s Freedom or Genesis platforms for robotics and liquid handling. The result was the identification of protein interaction networks, providing insight into the constituents of the various cellular machines, many of which would turn out to be involved in disease.

The two-hybrid approach is limited to those interactions that can form within the context of the yeast nucleus, and the assay is known to suffer from both false positive and false negative interactions. A complementary methodology is based on biochemical purification of protein complexes and their analysis by mass spectrometry (see Dec. 2004 Bio•IT World, page 18).

According to two-hybrid inventor Fields, the integration of the results of both approaches has been shown to be a strong predictor of protein function, as has the integration of results of such analyses across species (see PNAS 102, 1974-79; 2005). In the latter report, Trey Ideker and colleagues developed a bioinformatics tool called PathBLAST that enabled them to identify proteins across species involved in similar interaction networks from the Database of Interacting Proteins. They were able to predict a host of new interactions “that would not have been predicted from sequence similarity alone.”

Stacking The Chips
Numerous platforms exist for purification of proteins for mass spec analysis, ranging from immunoprecipitation to those based on solid-state chemistry and protein chips. An example of an integrated chip-based platform is Ciphergen’s ProteinChip System, Series 4000, introduced in mid-2004. This system includes Ciphergen’s ProteinChip arrays, SELDI-TOF mass spec detection, and software called Biomarker Patterns for data analysis. Ciphergen offers a variety of different chromatographic chips for protein binding, and has recently introduced the SEND ProteinChip Array, specifically designed for peptide mass fingerprinting or SELDI mass spec sequencing.

In early 2005, Biacore introduced the T100 system for protein-interaction analysis. Biacore’s systems make use of changes in surface plasmon resonance energy to define characteristics of proteins in terms of their interaction with other molecules — including binding and dissociation rates and absolute affinities — in real time.

The new release is a milestone in bringing this technology to the world of drug development. The system is compliant with FDA data security demands, and David Myszka of the University of Utah’s Center for Biomolecular Interaction Analysis, calls  it the “next generation” of SPR-based instrumentation.

The fundamental technology is similar to that of earlier instruments, but the user interface and data analysis programs have been significantly improved, making assay setup and analysis much more intuitive and user-friendly. Key applications include high-resolution immunological analyses and characterization of drug-protein interactions. 

Robert M. Frederickson is a biotech writer based in Seattle. E-mail:

Click here to login and leave a comment.  


  • Avatar

    Building protein interaction networks is indeed vital for continued discoveries in the biotechnology and academic molecular biology. The technology that I helped discover uses the static protein interaction networks discovered by the technologies described above. The technology ( then uses the dynamics of those networks to predict cancer patient outcome. What's interesting is that we have shown in our proof of principal study (Taylor et al, Feb 2009, Nature Biotech) that the predictive power of the test is limited by the size of the available protein interaction network. We are hoping that continued study of the human protein interaction network will allow us to better predict using this technology.

Add Comment

Text Only 2000 character limit

Page 1 of 1

For reprints and/or copyright permission, please contact  Jay Mulhern, (781) 972-1359,