From: LISA
Date: 15 February 2012 08:06
Hi all,
I have a DNA binding protein. I can get crystals when I mix 8-28 nt dsDNA with my protein. But neither of them has good diffraction. Some biochemical data said the longer of DNA, the tigher of the binding betwwen DNA and my protein. The binding is not sequence-specfic. Does anyone have suggestion of the optimization? What is the good length of DNA for crystallization?
Thank you.
Lisa
----------
From: Antony Oliver
Lisa, there isn't unfortunately a hard and fast rule for the length of DNA used in co-crystallisation. It usually is just a case of screening different lengths, permuting the sequence, and investigating overhangs or gaps in the DNA duplex. We generally work with oligos between 8 and 21 nts in length, but there are many examples of longer DNAs being co-crystallised, the nucleosome comes to mind as an extreme example.
Do you know if the protein binds better to DNA containing secondary structure elements, such as hairpin loops? This can make a difference, especially when you don't have sequence-specificity.
Tony O.
---
Mobile Account
---
----------
From: James Stroud
Use 5' overhangs of two and make the DNA 10, 11, 15, 20, 21 25, 26, 30, or 31 bases in length. Count the overhangs in the length.
If you don't know where to start, try 15, 25, and 26 first because they will make 2(1) screws, which are good for crystals.
James
----------
From: William G. Scott
This paper is a favorite: http://www.ncbi.nlm.nih.gov/pubmed/2160019
J Mol Biol. 1990 May 5;213(1):159-66.
Crystallization of Escherichia coli catabolite gene activator protein with its DNA binding site. The use of modular DNA.
----------
From: Xun Lu
Hi Lisa,
I will second James' suggestion. DNA packing seems really important, and making the DNA length as X half turns is usually good for packing (X=2, 3, 4 ...).
Another thing you might want to try is Hoogstein base pairing.
Cheers,
Xun
--
Date: 15 February 2012 08:06
Hi all,
I have a DNA binding protein. I can get crystals when I mix 8-28 nt dsDNA with my protein. But neither of them has good diffraction. Some biochemical data said the longer of DNA, the tigher of the binding betwwen DNA and my protein. The binding is not sequence-specfic. Does anyone have suggestion of the optimization? What is the good length of DNA for crystallization?
Thank you.
Lisa
----------
From: Antony Oliver
Lisa, there isn't unfortunately a hard and fast rule for the length of DNA used in co-crystallisation. It usually is just a case of screening different lengths, permuting the sequence, and investigating overhangs or gaps in the DNA duplex. We generally work with oligos between 8 and 21 nts in length, but there are many examples of longer DNAs being co-crystallised, the nucleosome comes to mind as an extreme example.
Do you know if the protein binds better to DNA containing secondary structure elements, such as hairpin loops? This can make a difference, especially when you don't have sequence-specificity.
Tony O.
---
Mobile Account
---
----------
From: James Stroud
Use 5' overhangs of two and make the DNA 10, 11, 15, 20, 21 25, 26, 30, or 31 bases in length. Count the overhangs in the length.
If you don't know where to start, try 15, 25, and 26 first because they will make 2(1) screws, which are good for crystals.
James
----------
From: William G. Scott
This paper is a favorite: http://www.ncbi.nlm.nih.gov/pubmed/2160019
J Mol Biol. 1990 May 5;213(1):159-66.
Crystallization of Escherichia coli catabolite gene activator protein with its DNA binding site. The use of modular DNA.
----------
From: Xun Lu
Hi Lisa,
I will second James' suggestion. DNA packing seems really important, and making the DNA length as X half turns is usually good for packing (X=2, 3, 4 ...).
Another thing you might want to try is Hoogstein base pairing.
Cheers,
Xun
No comments:
Post a Comment