In the opening scene of the film Twin Peaks, directed by David Lynch, a black and white TV screen flickers painfully, intrusively, meaninglessly. When an ax falls on the screen, exploding it to the accompaniment of a bloodcurdling scream, it's a relief. This is because our brains prefer to see clearly outlined images that can be construed into meaningful objects.
The televised image is really a code – it's based on a table for matching the coordinates of a pixel with a particular color. This type of "conversion table" offers an analogy for the genetic code, which can be seen as a kind of spread sheet for matching codons (three-letter sequences of DNA) to corresponding amino acids (the building blocks of proteins). We find the flickering TV screen disturbing because it contains no code: There is no relation between the color of a pixel and its location, so our brains can't "make sense" of it. A code, therefore, is born at the exact instant in which such connections appear, allowing one type of information – location or DNA sequence – to be converted to another – color or amino acid.
Dr. Tsvi Tlusty of the Physics of Complex Systems Department imagines the primeval world was something like the flickering TV screen: Proteins were random assemblies of amino acids, with no genetic code to guide their construction. "If this is so," he says, "it should be possible to write a script – a mathematical model – to describe the birth of the genetic code from the meaningless 'flickering screen' of random proteins." What causes the clear forms encoded in the genes to emerge from the morass? Such a change, says Tlusty, takes place when the benefit is clear and the cost not too high. The genetic code gives an organism a significant advantage in that it allows the plans for a variety of proteins to be stored and reproduced as needed. The cost of the code is the investment in molecular machinery to decode the information and translate it into proteins. An economic-style analysis can identify the point at which cost and benefit reach a balance and it becomes advantageous to adopt a system of codes.
In his mathematical analysis, Tlusty found that the emergence of a code, whether it be an image on a TV screen or a molecular code in living cells, bears a strong resemblance to transitions in the world of materials. Thus, for instance, the changeover to encoded proteins can be likened to the transition of a material from a liquid to a gas. By this analogy, the young code was "smooth." On the screen, this means that abutting pixels are likely to have a similar color; in proteins, two similar codons will code for the same, or chemically similar, amino acids.
Tlusty's analysis showed that the picture shaped by the code as it comes into existence is tied to a mathematical problem known as the "four-color problem." This mathematical theorem describes, for instance, the upper limit in the number of amino acids. The mathematical "script" for the birth of a code appeared recently in the Journal of Theoretical Biology and will soon be published in Physical Review Letters.
Dr. Tsvi Tlusty's research is supported by the Clore Center for Biological Physics; the Asher and Jeannette Alhadeff Research Award; and the Philip M. Klutznick Fund for Research.
An Inexact Match
For a living cell to function, its molecules must, while swimming in the cell's thick, erratic molecular stew, pinpoint and then bind to specific counterparts – something like finding a friend in a Tokyo subway station during rush hour.
In the classical view of molecular recognition, the binding molecules fit each other like a lock and key. In reality, however, the key is often not an exact fit for the molecular lock, and such molecules need to deform in order to bind. Why would evolution choose such an inexact system?
The work of Dr. Tsvi Tlusty and research student Yonatan Savir of the Weizmann Institute's Physics of Complex Systems Department suggests a possible answer. They developed a simple biophysical model which indicates that in picking out the target molecule from a crowd of look-alikes, the recognizer has an advantage if the target is a slight mismatch. This may appear to be counterintuitive – why search for a key that does not match its lock exactly and then require that imperfect key to warp its shape to fit the lock?
The researchers' model shows that the key's deformation actually helps in discerning its locking counterpart. Although the energy required to deform the molecular key slightly lowers the probability of its binding to the right target, it also reduces by quite a bit the probability that it will bind to a wrong one. Thus the quality of recognition – i.e., the ratio of right to wrong binding probabilities – increases.
This so-called "conformational proofreading" may turn out to be a crucial factor affecting the evolution of biological systems, and it may also be useful in the design of artificial molecular recognition systems.