A subtle tool to study mankind’s diseases

One of the most powerful tools in today’s biological and medical science is the ability to artificially remove and add bits of DNA to an organism’s genome. This has helped scientists to understand problems caused by defective genes, for example, which have now been linked to thousands of human diseases. So far the technology has been limited to small segments of DNA. But four years ago, Francis Stewart and his colleagues at the European Molecular Biology Laboratory (Heidelberg) developed a new technique to engineer greater stretches of DNA in bacteria. The researchers, now working at the Biotec-Technical University in Dresden, have just used this method to engineer a complex set of changes in a mouse gene, in hopes of shedding light on human leukemias. Their work appears in the current edition of the journal Nature Biotechnology.

Over two decades ago, researchers learned to use bacteria as “copy machines” for DNA taken from other organisms. This was a huge step for biotechnology, because most types of research require billions of copies of a molecule under investigation. However, there was a limitation: researchers need to change the DNA molecules in precise ways and for large molecules, such as whole genes, this was tremendously difficult.

Stewart and his colleagues thought that bacteria could be taught to do better, so they “borrowed” a strategy that organisms such as mice and yeast use to repair breaks in DNA. Proteins called recombinases circulate through their cells, looking for loose DNA fragments that have familiar sequences.

“Recombinases assume that the fragments have been cut out of the DNA by mistake, so they try to glue them back into the genome in the right place,” says Giuseppe Testa, who headed the current study. “Sometimes they’re a bit over-industrious; they put in pieces that look right, such as variations of a gene that have been put into the cell by a researcher.”

Called homologous recombination, this process works a bit like a “find-and-replace” command in your word processor. Imagine you have typed “Stephen Q. Gould” everywhere, and suddenly discover that the middle initial should be “J”. The computer can be told to look for “Stephen” and “Gould” and replace what comes between them. In the same way, recombinases find recognizable sequences of DNA to the left and right of a target and replace what comes in between with the new sequence.

Homologous recombination was known to occur in bacteria, but it hadn’t been possible to use it to engineer DNA, as was the case in yeast and mouse stem cells. Stewart?s team decided to try to find a strain that could do it. “We ordered as many types of E. coli as we could, looking for defects in the way they repair their DNA,” he says. “After five months of work, Youming Zhang, a postdoc in the lab, found the strain.”

The group quickly identified the bacterial factors involved and turned them into a new tool called Red/ET recombination that is now being adopted by biologists all over the world. It’s one of the mainstays of Gene Bridges GmbH, a company that Stewart and his colleagues founded with EMBL to develop the commercial implications of the breakthrough.

“We have been pushing it to work with larger and larger bits of DNA,” Testa says, “and our latest project has been to engineer an entire artificial chromosome in bacteria. We’ve constructed a large, complex ’cassette’ that we’ve now inserted into a mouse in place of its normal gene.”

The gene that they chose is called mixed-lineage leukemia (Mll), and is known to become defective in childhood leukemias in humans. By inserting the artificial version into the mouse, researchers hope to understand how the defects lead to disease. “There are many things that can go wrong in this gene,” Testa says, “and we wanted to construct a version of it that would allow us to test as many aspects of the problem as possible.”

The artificial Mll that they have put into the mouse will permit a variety of experiments. It contains two defects in the genetic sequence that have been linked to leukemias. The cassette also contains control switches that allow each defect to be “switched on” whenever the researchers choose; they can also be left off. “We can study each mutation independently, or watch how they act together, or control the time at which each one acts,” Testa says. “This will give us a new look at subtle relationships between multiple defects.”

Many diseases are linked to single mutations; however, disease susceptibility also often relies on other sequence variations, known as polymorphisms, in the human population. “The Mll cassette shows, in principle, a simple way to study both a mutation and a related polymorphism in the gene of interest,” Testa says. “This aspect of making mouse models will become increasingly more important for authentic modeling of human disease susceptibility and the way organisms respond to drugs and we think that our work shows the way to set up these models”.

The new work also heralds a new era for genomic engineering in many living systems. “The Mll cassette is a first demonstration of what can be done with large DNA molecules,” Stewart says. “Red/ET recombination increases the size of DNA that can be comfortably engineered by more than ten times and opens up new possibilities for genomic engineering that will filter into standard practice in the next few years.”