How to improve protein solubility in E. coli? - (Aug/14/2011 )
I have been expressed my recombinant protein in E. coli and have tried all parameter as below:
System: pET-32b (trx tag and his tag at both C and N terminals)
Target protein: 72kDa (predicted to have 9 disulfide bonds by homology modeling, 5 by "disulfind bond prediction" and 11 by DiANNA prediction)
IPTG: from 0.01mM up to 1mM
Unfortunately, they are all overexpressed in the insoluble form...any idea to increase protein solubility? I think the main problem here is the no of disulfide bond.
Hola, first there are some prediction solubility tools (http://www.biotech.ou.edu/ http://mips.helmholtz-muenchen.de/proso/proso.seam ). Second there are some tricks to improve solubility that rarely works but it¨s necessary try if the prediction of solubility is high. induction at low temperature buffer with pH away from of isoelectric point, a bit of reducer in the lysys buffer, hipertonic expression media. and you can try with one periplasmic vector. but in my experience these size proteins are difficult to express soluble in bacteria system. Think in an alternative as yeast or insect cells. Wait for more opinions. Buena suerte
Hi astrowoon,
Protein solubility is a bit of black magic. There are multiple factors that influence it and multiple approaches to address such a problem.
1) Strain screen: You could check Origami, TUNER and Rosetta strains. Note on your IPTG concentration screening - if you are not using a special cell strain like TUNER which has a mutation in the lac permease important for the active transport of IPTG in the cell you are unlikely to see an effect.
2) Expression temperature screening: The easiest experiment that you can do is expression at low temperature. Overnight expression at 20 or 18 degC are known to result in higher soluble protein yields due to decreasing the rate of folding. Expressions as low as 12 degC overweekend have been used (I believe in this case might be necessary to use a particular E. coli strain as well). Another relatively easy thing to try is screening with strains of overexpressed chaperones (commonly GroEL/GroES,
3) Construct screening: If you are pursuing structural characterization of this protein you might be interested in screening different constructs of domains within this protien.
4) Tag screening: From the solubility enhancers, the MBP-tag is well known to promote solubility despite having the so-called passenger solubility effect where the target protein would appear soluble only because it is attached to a very soluble protein and when tag is removed the target aggregates.
5) Refolding: You say that your target protein is produced in insoluble form i.e. you observe inclusion bodies, is that correct? In such a case you could try refolding from purified inclusion bodies.
6) Alternative expression systems: Finally, another approach which might be more time consuming is expressing your protein in another expression system like insect or even mammalian cells. What is the native organism of this protein? Viral and mammalian proteins are infamous for overexpression in e. coli especially in their full length.
Hope that helps. I could provide you with details on each of the points above if you are interested in further information.
Thanks to Protolder and Prodes:) My protein is a cellobiohydrolase (a type of cellulase) from A. niger. My project is to improve the thermostability and enzyme activity of that enzyme via directed evolution. Therefore, it is unlikely I will continue with protein refolding as this is too troublesome to screen for improved variants (ie I need to refold every batch of putative protein variants before I could get functional/soluble enzyme to screen for).
I used a Trx tag as it promotes the formation of disulfide bond in E. coli cytoplasm. Together with the Origami strain, which has the thioredoxin reductase and gluthathione reductase mutations, the formation of disulfide bond should be much possible. Also, Origami has a mutation in lac permease. However, I saw no difference in the expression profile ranging from 0.01mM IPTG to1.0mM IPTG.
I had tried expression at 20oC, but it is all produced as inclusion bodies. By using different protein solubility prediction tools, I have got different results. Some say yes, and some not. Considering directed evolution is my next step, changing a host is a pain. E. coli offers the simplest way (though not as easy) compared to other hosts. Maybe I will try out what Protolder suggests, trying to optimise the media before proceed to the last resort: change a host.
Kindly drop me your opinions. Thank you so much!
@Prodes,
Could you tell me further regarding strains of E. coli with overexpressed chaperones? What strain will you suggest for my case? Thanks!
Hola,two more oppinions. Before form inclusion bodies protein would be as soluble form at low concentration, so if you have done a time course expression probably you will find it at short time and low temperature but at low concentration. An alternative are the periplasmic vectors where processed protein is isolated from the periplasmic space by osmotic shock remaining high concentration of non processed protein in citoplasm because processing machinery is saturated. These signals are in pET 12, 20b 22b, 25b 0r 27band more developed pET 39 and 40 with an enviroment adequated for S-S isomerization.If you add a bit of glucose at the medium you could have a bit or repression (not always), and with glucose lactose mix in lacY+ strains you coud have an effect as overnight autoinduction system of Novagen. Check the on line manual of pET vectors particullary these periplasmic ones by your convienence as to do a try. Buena suerte
Try another tag like GST or MBP. Especially MBP improves solubility.
astrowoon on Wed Aug 17 02:33:34 2011 said:
@Prodes,
Could you tell me further regarding strains of E. coli with overexpressed chaperones? What strain will you suggest for my case? Thanks!
HI,
You are in similar waters
I am also working on protein engineering using directed evolution techniques.My experience was with strains described in this paper:
I would seriously consider working with yeast though! You say that originally the protein comes from A. niger. Therefore, due to the fact that it probably requires correct disulfide bonds for its folding and maybe even some glycosylation pattern I would take benefit of working with since they are also quite manageable for various high-throughput screening methods. For example, if you have plan a colony based screen it is likely to work in yeast as well since they form colonies too.
Lastly, if I understand correctly there is not available structure for this protein but is there a good template for homology modeling? If the answer is yes, you could probably obtain a very good model which would give you a better idea of how many disulfide bonds there are and the glycosylation pattern which all might affect the protein's solubility.
Cheers
Thanks all...and Prodes, your suggested journal helps me a lot!
I also have a large hydrophobic protein that I use this combination - NEB SHuttle Express T7 + TaKaRa Chaperone plasmid + pCold I - Does a nice job.
Btw, the strain that you use 12C for expression is Arctic Express.