Engineering better medicine. Jim Collins is well known for applying engineering tools to biological problems and arriving at creative and effective solutions. Now the ENG biomedical engineering professor and codirector of the Center for Biodynamics and a team including Timothy Gardner, an ENG biomedical engineering assistant professor, graduate student David Lorenz (ENG’05), and Diego di Bernardo of the Telethon Institute of Genetics and Medicine in Naples, Italy, have developed a simple, direct way to understand how genes and proteins interact to regulate processes within a cell, a tissue, or an organism. Using this new approach they are able to map biological networks with a minimum of experimental measurements.

Their new development provides a powerful tool to better understand complex biological processes and how they can malfunction. It is of particular interest and value for pharmaceutical research, since it can predict how pharmacological compounds will affect cell processes.

Called network identification by multiple regression (NIR), the approach is similar to that used to engineer large communication and control systems. In this systems engineering method, the researchers systematically perturb a cell with a genetic or a chemical stimulus, and measure how the changes impact mRNA or protein concentrations. They then apply computational algorithms to derive the strength and direction of all other regulatory interactions within the network, creating a map of the myriad gene and protein interactions that are possible.

In recent experiments, the researchers’ model correctly identified the major regulatory genes and their targets in a nine-gene subnetwork of the bacteria E coli, called the SOS pathway. They demonstrated that the algorithms were robust in the presence of biological noise, and are scalable to large networks.

This work was reported in the July 4 issue of the journal Science.

Scanning for defects. After 10 years of research and more than 120 million, the single genetic defect that causes cystic fibrosis was identified in 1989. Although researchers have developed faster and less expensive methods to identify such disease-related genetic changes, or SNPs (single-nucleotide polymorphisms), for most diagnostic purposes it is crucial to be able to simultaneously identify problematic areas, which may not occur at the same location, on both copies of the inherited gene associated with a particular disease or dysfunction.

Chunming Ding, a research assistant professor at the Center for Advanced Biotechnology (CAB), and Charles Cantor, CAB director and an ENG professor of biomedical engineering, recently developed a new technique that can rapidly scan for defects at widely separated locations along the genome. The novel approach holds great promise for diagnosis, treatment, and counseling for people at risk for genetic disorders.

Known as M1-PCR, the technique can identify mutations in halotypes, genes located close together on a given region of a chromosome that are usually inherited together from one parent. Earlier attempts at halotyping were labor-intensive, limited to short lengths of DNA, and difficult to interpret without prior knowledge of an individual’s genetic profile. With M1-PCR, the researchers were able to significantly reduce the sample amount and the time needed for processing, increase the span of DNA that can be tested, and eliminate the necessity for pedigree data.

The technique relies on simultaneously amplifying a 100-base-pair region around each SNP, rather than trying to amplify the entire genomic region containing all the SNPs of interest. Using the highly automated genetic amplification and analysis system that Cantor developed for Sequenom, Inc., of San Diego, Calif., where he is chief scientific officer, the researchers were able to build accurate pictures of halotypes of potential disease genes along distances tenfold longer than those described by any existing method.

The M1-PCR research appeared in the Proceedings of the National Academy of Science on June 11.