Can Amino Acids Sing? An Australian scientist may have found how and why amino acids fold into certain sequences--leading to possible novel new drug treatments.  At Monash University in Melbourne, associate professor Irena Cosic has developed this unique approach to drug design.       Biochemical reactions in living organisms are all controlled by biological catalysts or regulators, and most of these are proteins.  It has long been known that proteins depend on their three-dimensional shape for their action.  That shape, in turn, depends on protein structure.
    Proteins are long chains made of sequentially linked simpler molecules known as amino acids.  This sequence ultimately determines the three-dimensional shape into which the protein will fold, as well as its activity.    But exactly how protein shape and function emerge from the sequencing of amino acids has been the subject of laborious research, which until now has produced no great leap forward.
    As an engineer, Dr. Cosic decided to approach the problem mathematically--she decided to represent every amino acid in the sequence with a number, as a measure of each amino acid's ability to interact to determine this measure, Dr. Cosic used an estimate of the energy of the most loosely bound electrons of each of the 20 naturally occurring amino acids. These electrons are at the heart of chemical interaction.
    Having represented a protein sequence by a string of numbers, she could then draw a graph of the number sequence of amino acids along a protein.  Not surprisingly, what emerged looked like a spiky squiggle.  But to an electrical engineer like Dr. Cosic, such squiggles looked like signal sand an invitation to investigate further.
    When electrical engineers analyze complex signals--such as the trace of a heartbeat or radio waves from outer space--they break them down into a series of regular wavelike curves which, when added together, recreate the original signal.  Each component wave has a frequency measured in hertz, the number of times the wave form is repeated each second.
    Dr. Cosic was able to take whole protein sequences and graph the contribution of each component wave form against its frequency.  She then compared the results for groups of proteins which performed a similar function and for proteins which performed different functions.
    What emerged was astounding.  Among their component waves, proteins which performed the same function--a group of growth factors, for example--all possessed one waveform in common, a single characteristic frequency which they shared.  In contrast, proteins which were unrelated in function shared no such frequency.  What is more, the characteristic frequencies were different for groups of proteins performing different functions.
    Even more astonishing was that when Dr. Cosic analyzed the target molecules which interact with proteins--the receptors to which a group of proteins bind, for instance--she found that the characteristic frequency of the protein group matched that of its target.
    "They recognize each other on the bias of frequency.  It's very obvious to an electrical engineer.  It's like tuning into a radio station."
    Dr. Cosic then explored the possibility that electromagnetic waves of the right frequency levels could be produced by electron movement along proteins.  When she calculated such movement theoretically, she found the energy frequencies were typical of visible light.  So she investigated groups of light-sensitive proteins, comparing the values she calculated for their characteristic frequencies with the frequency of the light ot which they were sensitive.  Sure enough, the values matched.
    But the ultimate test of any scientific hypothesis is prediction.  She reasoned that if her theory was correct, and the action of group of functionally similar proteins was related to their characteristic frequency, it should be possible to mimic that action by designing a sequence of amino acids with the same characteristic frequency.
    And this is just what Dr. Cosic has done.  The first group of proteins she chose was on the outside coating of the human immunodeficiency virus (HIV). After calculating the characteristic frequency of these proteins, she designed a totally unrelated amino acid sequence ot match that frequency.  The sequence was constructed and tested at the Universite de Lyon in France.  It was found that the newly designed amino acid sequence could trigger the immune system to recognize HIV.  As such, the sequence might be useful as the basis of a vaccine.
    Dr. Cosic, who came to Australia from her native Serbia in 1989, is currently deputy head of Monash's Electrical and Computer Systems Engineering department.  She is also the author of the recently published The Resonant Recognition Model of Micromolecular Bioactivity.