Unfolding The Genetic Code

[Guest guest]

Unfolding The Genetic Code

http://www.medicalnewstoday.com/medicalnews.php?newsid=58422

It turns out that sequencing the human genome - determining the

order of DNA building blocks -- has not completely cracked the code

of how DNA directs various cellular processes. In addition to the

sequence of the base pairs, the instructions are in the packaging -

how DNA is folded within a cell.

Virginia Tech researchers used novel methodology and the

university's System X supercomputer to carry out what is probably

the first simulation that explores full range of motions of a DNA

strand of 147 base pairs, the length that is required to form the

fundamental unit of DNA packing in the living cells -- the

nucleosome. Contrary to a long-held belief that DNA is hard to bend,

the simulation shows in crisp atomic detail that DNA is considerably

more flexible than commonly thought.

The research is published in the December issue of the Biophysical

Journal, in the article " A Computational Study of Nucleosomal DNA

Flexibility, " by Jory Zmuda Ruscio of Leesburg, Va., a Ph.D. student

in the Genetics, Bioinformatics and Computational Biology Program at

Virginia Tech, and ey Onufriev of Blacksburg, assistant

professor of computer sciences and physics at Virginia Tech. They

have been invited to do a platform presentation at the 51st

Biophysical Society Annual Meeting in Baltimore in March.

There is about 12 feet of DNA in a human cell but it is packaged

into nucleosomes - lengths of 147 base pairs each wrapped around

eight special proteins. A nucleosome looks kind of like the lumpy

beginning of a rubber-band ball. Or maybe more like a lumpy worm

coil. Uncoiled, the worm wiggles, flexes, and even kinks, according

to a simulation performed on System X.

As we know from watching forensic detective shows on TV, the DNA in

all of an individual's cells is identical. The DNA in fingernail

cells is exactly the same as in muscle. Yet the cells are

different. " This is because, roughly speaking, the DNA in different

cell types is packed differently and the complexes it forms with the

surrounding proteins are in different positions, so only the

relevant part of the code can be read at a time, " said

Onufriev. " Although nobody knows exactly how it happens, you can

imagine reading only what you can see on a part of a crumpled

newspaper. "

The traditional view is that DNA is relatively rigid and that

considerable energy is required when it needs to be bent to form

protein-DNA complexes. However, recent experiments (Nature, Aug. 17,

2006) have begun to challenge that view. " The famous double-helix

may be much more flexible than previously thought, " said Onufriev.

The Virginia Tech research responded to this debate. Using 128 of

System X's 1,100 processors, the research resulted in a System X

movie revealing DNA wiggling like a worm, showing greater

flexibility than expected from the traditional view. The DNA packing

in the nucleosome is also found to be surprisingly loose. " The

implication is that it may not cost much energy to bend the DNA -

even to bend sharply, " said Onufriev.

The methodology that is making it possible is based on the so-

called " implicit solvent " approach being developed by

Onufriev. " Biology does not happen in a vacuum, " he said. " We are 75

percent water, and the effect of the water environment must be taken

into account when studying biomolecules. "

Previous simulations were often slowed because they accounted for

the water that is present in living systems. For instance, in early

studies of protein folding, only a few percent of the computing

effort was being spent on the activity of the protein while the rest

accounted for the activity of the surrounding fluids. The " implicit

solvent " approach accounts for the role of water on average, but the

movements of individual water molecules are not predicted, freeing

computation capacity for simulation of whatever protein is being

studied.

" Experiment cannot always probe atomic detail of living molecules

because they are too small and often move too fast, said

Onufriev. " But we can combine computational power with good

algorithms to simulate these motions at high (atom-scale)

resolution.

" It is an exciting time to do molecular modeling, " he said. " The

computing power and the methodology have come to the point that we

can begin to fully probe biology on timescales very relevant to

living things - such as DNA packing. "

Virginia Tech's System X supercomputer was critical to this

research, he said. " It was the combination of its sheer compute

power with the algorithmic advantages that made it possible to run

molecular simulations on that scale. "

So far, the Virginia Tech research team addressed the question of

how flexible the DNA is, which is only a small piece of the " second

part of the genetic code " puzzle, Onufriev said. " However, this

small piece should pave the way to addressing bigger questions, such

as 'Exactly how is the tightly packed genetic content read by

cellular machines " ' "

" Atomic level simulations can complement experimentation and narrow

competing theories, " said Onufriev. " For systems as large as the

nucleosome, simulations using virtual water may be the only

practical way to estimate the stability of various confirmations, "

he said.

How DNA bends and flexes is critical for many cellular processes

including cell differentiation and DNA replication. Although also

observed in recent experiments, this unusual DNA flexibility is

still unexplained. " Now seeing that DNA is not as hard to bend may

lead to radical changes in our perspective, " said Onufriev. " We are

using these detailed pictures to see exactly how DNA bends and to

understand the details of the mechanism behind it, something that is

very hard or impossible to do experimentally. "

Onufriev and his group of biochemistry, physics, biology, and other

computer science researchers received a $1.1 million grant from the

National Institutes of Health to develop high performance computing

methodology to create molecular models and to probe in atomic detail

the mechanisms of biology.

The purpose of the NIH award is to develop the methodology for

computer simulations of complex biological processes and address the

question of the atomic mechanism of DNA flexibility, Onufriev

said. " This research may not only provide fundamental insights into

how life works at the molecular level, but also has applications in

drug discovery and in particular for rational drug design, which is

an important consideration for the NIH. "