Model of possible interaction of the (gene regulating) Cro represser protein
and B-DNA (Deoxyribonucleic Acid). Work done in 1981?? by Dr. Robert Langridge.
This demo can be run by typing: midas -f crodna (on a DEC Alpha, replace '-f' with '-fullscreen')
In Midas, type: source demo.script
The demo starts with a picture of B-DNA on the left and the Cro protein
on the right.
Next the cro protein is shown to be a dimer by linking together the main
carbon atom in the constituent amino acids of the protein. This representation
gives a feel for the structure of the protein by showing the helical and
strand portions of the protein.
Next the image is rotated to show off the fit of the cro protein with DNA.
Then a ball and stick image of DNA is shown with a ribbon diagram of the
protein. The ribbon accents the helical and strand portions of the protein
even more but it is drawn too slowly to be rotated interactively. Click
on the mouse to continue the demo.
Next a space-filling image, otherwise known as a Corey- Pauling-Koltun (CPK)
style image of the molecule is shown with shadows and specular highlights.
The shadows and highlights add to the three dimensional look of the image
but are computationally expensive and thus not suitable when rotating interactively.
Click on the mouse to continue the demo.
Next we zoom in on a possible bonding site between cro and DNA. The rest
of the atoms in the amino acid near DNA is shown and the atoms are labelled
and colored by atom type. The dashed line shows where a distance is being
measured between a nitrogen atom in DNA and one in the protein. The distance
is 2.51 Angstroms, which is a little close for hydrogen bonding (3 Angstroms
would be better) but indicates that it might be possible with a few bond
rotations. The dot surface is at the bonding distance around the atoms and
the slight overlap between the hydrogen in the protein and the nitrogen
in DNA improves the likelihood of hydrogen bonding.
Next, the DNA molecule is mutated by substituting a different base, and
the visible overlap indicates that the mutation wouldn't work. Then, the
mutation in undone, and the base above the previous base is colored. Two
distances and two angles are monitored and four bond rotations are set up
in protein. Again we are looking for possible hydrogen bonding, this time
between two bases and the protein. The protein's bonds are rotated and then
the picture is rotated to show the result. The amount of surface overlap
looks good for hydrogen bonding.
This is enough evidence of bonding between the Cro protein and DNA that
wet lab experiments might be done to confirmIn general, the computer-based modelling helps to reduce the number of experiments
that need to be done by a chemist.
Epilogue: Later studies show that Cro doesn't interact with DNA in the way
that was shown but probably did during its evolution. (Who??)