A few ideas:

1) go even higher with protein concentration. Some ultra-soluble proteins may 
need >100 mg/mL. You may have to set up drops at higher ratios of 
protein:precipitant to achieve this. e.g. 3:1 or more. Use a test such as the 
Hampton PCT to tell you when you are in the right concentration range.

2) reductive methylation of lysine residues as Stephen suggested. This often 
reduces the solubility of proteins and will change the surface properties.

3) get rid of any glycosylation as the surface sugars increase solubility (and 
are flexible anyway and so bad for crystallization)

4) check the few drops which have precipitate and look for common factor e.g. 
>2M Ammonium Sulphate. You could then add a small amount of this ingredient 
(e.g. 100mM Ammonium Sulphate) to your protein and then re-screen.

5) Wait a bit longer. One week is still early days.

Good luck
Tom

**                          Tom Walter B.Sc. M.Res.                   **
** Oxford Protein Production Facility        Tel: +44 (0)1865 287747  **
** Wellcome Trust Centre for Human Genetics  Fax: +44 (0)1865 287547  **
** Roosevelt Drive                           [EMAIL PROTECTED]   **
** Headington, Oxford OX3 7BN                http://www.oppf.ox.ac.uk **


---- Original message ----
>Date: Tue, 13 May 2008 11:16:41 -0500
>From: Jennifer Han-Chun Tsai <[EMAIL PROTECTED]>  
>Subject: [ccp4bb] problem of crystallization  
>To: [email protected]
>
>   Hi,
>    
>       This topic is not related to CCP4. I am having problem of
>   crystallizing one protein. It's a pretty small protein with
>   size around 15kDa. I have stock concentration around 100mg/mL.
>   Crystallization plates I set up are with concentration of
>   10mg/mL, 30mg/mL and 50mg/mL. All the plates had been set up
>   at least one week. Only around 5 wells per plate or less
>   formed precipitation. The rest of wells are pretty clear
>   still. Is there any suggestion for reducing protein solubility
>   or increasing the chance of getting crystals?
>    
>   Thanks for your time,
>   Jennifer 

Reply via email to