Development of chips that mimic living cells

A single cell in the human body is about ten thousand times more energetically efficient than any digital nano-transistor that exists today and which constitutes the basic building block in electronic chips. In one second, a biological cell carries out about ten million chemical reactions consuming energy in the cumulative amount of one picowatt

A single cell in the human body is about ten thousand times more energetically efficient than any digital nano-transistor that exists today and which constitutes the basic building block in electronic chips. In one second, a biological cell performs about ten million chemical reactions that consume energy in the cumulative amount of one picowatt.

Scientist Rahul Sarpeshkar of the MIT Research Institute is now applying architectural principles taken from these energy-efficient cells to the design of hybrid digital-analog electronic circuits, energy-efficient. Such circuits could be used one day in the future for the development of extremely fast supercomputers that would enable the prediction of complex reactions of cells to drugs. They may also help researchers develop artificial genetic circuits in cells.

In his new book, "Ultra Low Power Bioelectronics" (Cambridge University Press, 2010), the researcher points out the many similarities between chemical reactions occurring in a cell and between the flow of electricity in a parallel electrical circuit. It describes how biological cells perform reliable calculations despite background noise and sensitive components found in them (referring to random changes in signals - both electrical and genetic). Circuits that will be built in the future based on these principles will overcome these noises while maintaining the high energy efficiency. Promising applications in this area include image processors in mobile phones or special brain implants for the blind.

"Electric circuits are a language for representing and trying to understand almost anything in the world, whether it's networks in biology, or whether it's vehicles," says the researcher, a professor of electronic engineering and computer science at MIT. "There is a uniform and efficient way to look at the biological world using electric circuits."

Electrical circuit engineers are already familiar with hundreds of approaches for operating analog circuits at low energy, for amplifying signals and reducing background noise, knowledge that has helped them develop energy-efficient electronic devices, such as telephones and portable computers, music players, and the like.

"Here is a field in which about fifty years have been invested in understanding the design of complex systems," the researcher notes, referring to the field of electrical engineering. "Now we can think about biology in the same way." He hopes that physicists, engineers, biologists and biotechnologists will work together as pioneers in this new field, which he calls "cytomorphic" electronics (cytomorphic, whose design was inspired by biological cells).

The researcher, an electronics engineer with many years of experience in developing energy-efficient bio-electronic circuits, often pondered how to exploit the connections between biology and electronics. In 2009, he developed an energy-efficient radio chip that mimics the structure of the human cochlea to filter and process mobile phone, internet, radio and television signals, faster and more energy-efficient than could be believed at the time.

This chip, known as a radio frequency (RF) earlobe, is an example of "neuromorphic electronic components", a field that has existed for about twenty years and was founded by the scientist Carver Mead who was the researcher's mentor at Caltech. Neuromorphic circuits mimic biological structures found in the nervous system, such as the cochlea, retina and brain cells. The researcher's conceptual expansion from the field of neuromorphic electronics to cytomorphic electronics is based on his examination of the questions underlying the dynamics of chemical reactions and the flow of electrons within circuits parallel to them. He discovered that these questions, which help predict the behavior of chemical reactions, are surprisingly similar, even in their background noise properties.

Chemical reactions (for example - the acceptance of water from hydrogen and oxygen) occur at a practical rate only if enough energy is invested in them that reduces the energy barrier that prevents them from occurring normally. A catalyst, such as an enzyme, is able to reduce these barriers. Similarly, electrons flow within an electrical circuit when a transistor utilizes external voltage energy to allow them to reduce the barriers preventing them from moving from the source of the transistor to its exit point. Changes in the voltage energy supplied to the transistor lowers the barrier and increases the flow of electrons in the transistors, just as enzymes are added to a chemical reaction to speed it up.

Ultimately, biological cells can be thought of as circuits that use particles, ions, proteins, and DNA, instead of electrons and transistors. This acceptance implies that it may and will be possible to develop electronic chips - which the researcher calls "cellular chemical computers" - that mimic chemical reactions efficiently and on a very fast time scale.

An example of a possible powerful application of such a circuit lies in the visualization of a genetic network - the interrelationships between genes and proteins that control the activity and fate of the living cell. In his previous article, from 2009, the researcher demonstrated how such a circuit is able to provide visualization of any genetic network using a chip. For example, these circuits can simulate the interrelationships between genes involved in the breakdown of sugar (lactose) and between transcription factors that regulate expressions in bacterial cells.

In the next step, the researcher plans to develop circuits that mimic the interactions within the entire cellular genome, relationships that are important to scientists in that they allow them to better understand and treat complex diseases such as cancer and diabetes. In the end, researchers may be able to use such chips to simulate the entire human body, the researcher believes. These chips will be much faster than computers that provide simulations today, which are very limited in simulating the effects of non-linear background noise that exists inside the cell. He also examines how the design principles of these circuits can help in the genetic engineering of cells so that they perform useful functions, for example - sensitive and fast detection of toxins and poisons in the environment.

The news from the research institute