[ccp4bb]: summary of archeal protein expression in Ecoli

[Narayanan Manoj <nm77@cornell.edu>]

***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***

Question

>           I am trying to express an archeal protein in Ecoli. Can
> someone name some references that lists commonly used expression and
> purification protocols in such cases.

Replies

1)  We have expressed a number of archeal proteins in E. coli - AFAIK
there
are no special methods for that particular group, provided that you make
sure that the codon usage is optimized for high expression. The usual
problems pertaining to expression of toxic proteins, DNA-binding
proteins etc. apply in full.
As far as purification is concerned - it is tempting to purify by heat
treatment if your protein is thermostable, however in some cases heat
treatment results in strange effects e.g. aggregation. Thus, we mostly
use whatever methods seem to be working. If heat treatment works -
great, if it does not - no biggie.

--Artem G. Evdokimov

P.S. surprisingly, thermostable selenomethionine-labeled proteins seem
to survive heat treatment without too much trouble - with the following
caveat: exposed selenomethionines are frequently damaged, buried SeMEt
residues seem to survive just fine (tested with 70 and 90 C incubations
10-30 minutes in length). MS indicates that the most common result of
heat-treatment of exposed SeMet residues is the loss of Me-Se- fragment
(likely resulting in formation of an alkene where the group used to be,
akin to a well known organic name reaction). Perhaps this can be used to
estimate the content of buried methionine in an unknown thermostable
protein (if one needs to know if there's enough for MAD, for example)
.....

2) I can provide you my own experience since i am dealing with
a protein from Sulfolobus solfataricus.
You should care about the codon bias problem. Try to use an additional
plasmid in you expressnion system that encodes for rare tRNAs
(see Stratagen Codon plus plasmid, there
is also some home-made thingies like the one from Sung-Hou Kim lab,
Berkeley)
We used with very much success special DE3 strains the so-called C41 and
C43
wich are special mutants of the BL21 (DE3) strain (John E Walker ,
MRC). We always got good expression in that for various proteins.
For the purification you should try an heat-denaturation step. 
Resuspend the 
cells in your buffer warm up, in a seperated tube some buffer at X
degrees
(much larger volume than you sample) and mix both in such a way
that you get in a step at a temperature wich corresponds to the
optimal temperature growth of the archaeon from which comes your
protein. Don't put
glycerol in your buffer, only a biological buffer and some salt. Don't
use tris
because it has a pKa dependant on the temperature.
Don't just place your lysate (in the buffer room temperature) in the hot
bath,
because you will induce a temperature gradiant in your sample tube
and proteases will get really active for some time.
Then do an ultra-speed centrifugation above 100 000 g, you should then
get
a very clean lysate. Then, one or two chromatography steps should
provide you
a very clean sample for crystallization.
 --Greg.

3)most archaeal proteins I had my hands on worked with standard
protocols
using a T7 expression system (like pET from Novagen, BL21 or my
preferres strain B834 which you can also use for SeMet labeling). Use
some affinity tag and clone a specific cleavage site in between in case
the tag
interferes with crystallization. A TEV site and a His tag are obvious
choices.
Try both N- and C-termini. With hyperthermostable proteins, the
advantage for purification is that
a heat step followed by a hydrophobic column if possible (to remove the
nucleic acids, which don't bind to the column) is a highly efficient
protocol. When using tags, these proteins frequently don't bind too
tightly to the IMAC column because the termini are often not well
accessible.

-- bjoern

4)I assume you have checked your sequence for rare codons (RIL) ?
If this is not the case you should find out how many rare codons are
present  in your sequence before cloning your gene into E. coli.
A high content of rare codons >5% will reduce the yield of your protein
significantly in E.coli (in my case 10% rare codons; 50 fold increase of
my protein between normal BL21 and Codon Plus Cells)
I can recommend the following strain from Stratagene BL21 Codon Plus
RIL. You also might want to check the Rosetta E. coli strain.
Assuming your archeon has a high temperature optimum you can use heat
shock as a first purification step with a diluted crude extract (10 ml /
g wetweight).

Juergen


5) On a related issue: you can also of course back-translate and make
the 
synthetic gene with E.coli codons. Any clues as of how much this costs 
these days and who does it ? I got one company for 3.50 Euro per bp. Too 
expensive for my taste.

moreover a rare codon can result to half of your protein to be C-term 
truncated and happily sticking to N-agarose - in case of N-term tags ...
We need a good protease for removing C-term tags .. all I have seen cut 
close to the C-term of the linkers and thus would leave too many 
residues if tried with C-term tags ... Anything like that out there ?

                Tassos

Thanks a lot for all who replied
manoj

-- 
Narayanan Manoj, 
372, Baker Lab, 
Cornell University, Ithaca, NY 14853-1301
Ph: (607) 255 7517