Biologic Computer Designed to Kill Cancer Cells

This network recognizes certain cancer cells using logic combinations of five cancer-specific molecular factors, triggering cancer cells destruction.

Dr. Yaakov Benenson, professor of synthetic biology at ETH Zurich (Switzerland), has spent a large part of his career developing biologic computers that operate in living cells. His goal is to construct biocomputers that detect molecules carrying important information about cell wellbeing and process this information to direct appropriate therapeutic response if the cell is found to be abnormal. Now, together with the Massachusetts Institute of Technology (MIT; Cambridge, MA, USA) professor Ron Weiss and a team of scientists including post-doctoral scholars Zhen Xie and Liliana Wroblewska, and a doctoral student Laura Prochazka, they made a major step towards reaching this goal.

In a study that has just been published in the September 2, 2011, issue of Science, they described a multigene synthetic “circuit” whose task is to distinguish between cancer and healthy cells and consequently target cancer cells for destruction. This circuit works by sampling and integrating five intracellular cancer-specific molecular factors and their concentration. The circuit makes a positive identification only when all factors are present in the cell, resulting in very precise cancer detection. Researchers hope that it can serve a basis for very specific anticancer treatments.

The scientists tested the gene network in two types of cultured human cells: cervical cancer cells, called HeLa cells, and normal cells. When the genetic biocomputer was introduced into the different cell types, only HeLa cells but not the healthy ones were destroyed.

Considerable groundwork was required to achieve this result. First Dr. Benenson and his team had to find out which combinations of molecules are unique to HeLa cells. They looked among the molecules that belong to the class of compounds known as microRNA (miRNA) and identified one miRNA combination, or profile, that was typical of a HeLa cell but not any other healthy cell type.

Finding the profile was a tricky task. In the human body, there are about 250 different healthy cell types. Moreover, there are many variants of cancer cells, of which hundreds can be grown in the laboratory. Still greater is the diversity of miRNA: between 500-1,000 different species have been described in human cells. “Each cell type, healthy or diseased, has different miRNA molecules switched on or off,” noted Dr. Benenson.

Creating a miRNA profile is not unlike finding a set of symptoms to diagnose a disease effectively: “One symptom alone, such as fever, can never characterize a disease. The more information is available to a doctor, the more reliable becomes his diagnosis,” explained the professor, who came to ETH from Harvard University (Cambridge, MA, USA) one and a half years ago. The researchers have therefore sought after several factors that accurately differentiate HeLa cancer cells from all other healthy cells. It turned out that a combination of only five specific miRNAs, some present at high levels and some present at very low levels, is enough to identify a HeLa cell among all healthy cells.

“The miRNA factors are subjected to Boolean calculations in the very cell in which they are detected. The biocomputer combines the factors using logic operations such as AND and NOT, and only generates the required outcome, namely cell death, when the entire calculation with all the factors results in a logical true value,” said Dr. Benenson. Indeed, the researchers were able to demonstrate that the network works very effectively in living cells, correctly combining all the intracellular factors and giving the correct diagnosis. This, according to Dr. Benenson, represents a significant achievement in the field.

In a next phase, the scientists want to assess this cellular computation in an appropriate animal model, with the aim to build diagnostic and therapeutic tools in the future. This may sound like science fiction, but Dr. Benenson believes that this is feasible. However, there are still complicated problems to solve, for example, the delivery of foreign genes into a cell efficiently and safely. Such DNA delivery is currently quite challenging. In particular, this approach requires temporary rather than permanent introduction of foreign genes into the cells, but the currently available techniques, both viral and chemical, are not fully developed, and need to be improved.

“We are still very far from a fully functional treatment method for humans. This work, however, is an important first step that demonstrates feasibility of such a selective diagnostic method at a single cell level,” concluded Dr. Benenson.

Related Links:
ETH Zurich
Massachusetts Institute of Technology