Looking at Cells with a Computer Developer's Eye

March 21, 2006--The most sophisticated computing processor is not on your desktop, but in the cells inside your body, scientists revealed last week at a physics meeting.

Most of the things that a cell does seem simple enough to be included in a middle school textbook: producing proteins, packaging waste, taking in food and reproducing. But the mystery at the heart of the cell is the nucleus. Less than a century ago we didn't understand what was inside. Now we know the nucleus is filled with molecular strings we call DNA and that it serves as the memory bank of the cell, reproducing when needed and providing information to direct all the cell's processes.

But how do we understand what DNA does? Can a knot of molecules serve as the master hand of the cell's day-to-day operations? Or is the whole process more distributed and more interesting?

James Shapiro, a geneticist and an expert in the physics of molecular biology, sees the answer: the nucleus is an information center at the service of cellular computing networks, and DNA is the physical medium in which the data is stored. Shapiro presented evidence for his idea at the annual March Meeting of the American Physical Society, held in Baltimore.

"DNA has to do more than code for protein structure," Shapiro said. Rather than seeing the genome -- the collection of DNA within a cell -- as a library of instruction books to make the cell's parts, DNA acts more like your home computer's memory and hard drive connected to the Internet. DNA, working with other molecules, takes part in many functions that make the cells work: copying and transmitting data, proofreading the product, sending copies of the instructions to daughter cells, repairing and even restructuring the data when needed. DNA, it would seem, is a read-write storage medium, Shapiro deduces.

And like your computer, data to do all these tasks has to be organized properly. "DNA by itself doesn't do much and requires other molecules, and since it has to work other molecules, it requires formatting."

For example, DNA can be formatted to execute an algorithm, such as determining what action to take depending on what food is present.

The immune system is an example of the lengths cells can go to in restructuring their DNA for a specific function immune system cells must produce antibodies that are specifically structured to defend the body against invaders whose identity cannot be anticipated. With a finite amount of DNA, immune cells have to encode a virtually infinite variety of antigen-binding specificities in the right part of trillions of antibody molecule. They do this through a highly ordered but flexible series of DNA restructuring steps that rapidly evolve new protein sequences to recognize any conceivable invader. In genome restructuring, just as in genome expression and transmission, biological feedback circuits are an important part of the cell's information processing.

And rather than having one gene do one function, as scientists hypothesized before we knew about DNA and cellular networks, Shapiro says that genomes perform millions of tasks as part of a complex information processing system. Instead of looking at the components individually, Shapiro says, the 21st century way to look at the genome is as though it were part of an interactive information processing system, in some ways like a computer -- each component participating in many functions as the need arises.

Though most people may envision the nucleus as the "brain" of the cell, that analogy no longer holds. "Information processing involves molecules all over the cell," Shapiro says. "DNA is a key component of the process but cannot act on its own."

Looking at genomic information in the context of the cell as a whole can help us understand how different parts of the genome determine the relationships between species. Although species may share many of the same sequences coding for proteins, other kinds of information, repetitive DNA organizing genome structure, can be quite different, even in related organisms.

In addition, computer designers may glean important cues about high-throughput information processing by studying the layers of self-correction and efficiency that have been refined by millions of years of cell evolution. The cell processes hundreds of millions of biochemical and biomechanical events in each cycle -- and the error rate in all those processes is so low that the control processes keeping them all in synch far surpass what we can do now with computers, Shapiro said. "If we could reproduce that, we could be on to something."

Reference:

James A. Shapiro, A 21st Century View of Evolution: Genome System Architecture, Repetitive DNA, and Natural Genetic Engineering, Gene 345: 91-100

Contact:

Martha Heil
American Institute of Physics
Tel: 301-209-3088
mheil@aip.org