Expressing protein in a soluble form - What should be the next logical step? (Aug/02/2005 )
Howdy!
I'm in desparate need of advice regarding what I should do next. I have a cDNA cloned into pET-24a for expression of ~55kDa HIS-tagged protein in BL21DE3. I get fantastic expression, but in the pellet fraction indicating it is insoluble. I tried a time-course expression (@ 16 degrees) to see if at the earlier periods of induction (IPTG), I could get it in a soluble form. However, from 30 mins when its induction is obvious (!!), it is insoluble. Then I tried using 8M urea to solubilize. This time it did come up in the soluble fraction (although only ~ 50%). I am aware that 8M urea is in the higher end and the protein will have to be checked for functionality after refolding (function of my protein not clear, so chekcing for functionality may be useless? I suppose I can reduce conc. of urea to reduce the refolding problem, but that may further comprise the amount of protein in the soluble part.
I have asked around, and there are some suggestions as to what I could do:
1) try other vectors as the protein may be soluble when expressed by a different vector - my problem with this is that my protein is 50% hydrophobic (although it is a cytoplasmic transcription factor in mammalian cells), and I'm worried that the vector may not make a diff (what's your opinion?). Also, why would changing the vector change protein solubility when the tag is the same?
2) express the protein in other systems (insect cells, cell lines, yeast), which is possible but tough.
I would truly appreciate any input. Thanks.
ggUss
Just my opinions, but...
1) Choice of vector shouldn't make any difference, at least not for what you're describing. That generally only helps if you're having a problem with the expression being too rapid for the system to handle and you're forming inclusion bodies. So, if you think that's happening, you could try it, but since you expressed at 16 degrees, this is highly unlikely.
However, if you were to express the protein as a fusion protein with GST or MBP, you may be able to pull your protein out of the pellet by virtue of the fusion's high solubility. We've seen this many times. Of course, the real question at that point is whether you can deal with a fusion protein stuck on there. Sometimes if you cleave it off, your protein of interest just goes right back to the pellet fraction.
2) Can't really say, as I've had very little experience with non-bacterial expression systems. I can see it making a difference if your protein normally undergoes posttranslational modification, or if it's a membrane protein or something, but if it's simply compartment-specific you'll probably run into the same problems once the cells are lysed.
My suggestion - try a fusion construct. You can usually get away with a GST-tagged protein for most biochemical assays, given the appropriate control experiments, and you might solubilize your protein without the need for a renaturation step.
1) Choice of vector shouldn't make any difference, at least not for what you're describing. That generally only helps if you're having a problem with the expression being too rapid for the system to handle and you're forming inclusion bodies. So, if you think that's happening, you could try it, but since you expressed at 16 degrees, this is highly unlikely.
However, if you were to express the protein as a fusion protein with GST or MBP, you may be able to pull your protein out of the pellet by virtue of the fusion's high solubility. We've seen this many times. Of course, the real question at that point is whether you can deal with a fusion protein stuck on there. Sometimes if you cleave it off, your protein of interest just goes right back to the pellet fraction.
2) Can't really say, as I've had very little experience with non-bacterial expression systems. I can see it making a difference if your protein normally undergoes posttranslational modification, or if it's a membrane protein or something, but if it's simply compartment-specific you'll probably run into the same problems once the cells are lysed.
My suggestion - try a fusion construct. You can usually get away with a GST-tagged protein for most biochemical assays, given the appropriate control experiments, and you might solubilize your protein without the need for a renaturation step.
Thank you aludlam for your constructive insight. I will do try a fusion contruct as you have suggested since I have little faith in refolding the protein to its right conformation (whatever that is! the 3D structure is not even known!).
However, we were planning on using the purified protein for antibody generation and structure analysis. wouldn't the fusion hinder this process? if I haven't got a choice, which would be less immunogenic and obstructive- GST or MBP?
Thanks again for your time.
If you're doing structure work, you may want to look into other options. GST and MBP will both give you anomalous results for antibody generation. Occasionally fusion proteins can be crystallized for structure work, but it's obviously not the ideal situation.
If you have a peptide sequence, you could try looking at that with the ExPasy tools, particularly topology and secondary structure predictions. If you find a segment near a terminus that's highly hydrophobic or possibly transmembrane, you could try to cleave that off and express the construct. This would inhibit a full structural analysis, obviously, but it's a better option for antibody production if you can get a soluble truncation.
If you're left with no options for truncation, and a fusion construct is soluble, you could try to generate your antibodies with a fusion construct and then remove fusion-specific antibodies by affinity binding to GST/MBP alone. It would be labor-intensive, though, requiring lots of binding steps, and you may not be able to fully purify it away.
The other option is to make the fusion construct cleaveable and see if you can generate soluble protein after partial purification and removal of the fusion. It's tricky, but it can work in some circumstances.
If you have a peptide sequence, you could try looking at that with the ExPasy tools, particularly topology and secondary structure predictions. If you find a segment near a terminus that's highly hydrophobic or possibly transmembrane, you could try to cleave that off and express the construct. This would inhibit a full structural analysis, obviously, but it's a better option for antibody production if you can get a soluble truncation.
If you're left with no options for truncation, and a fusion construct is soluble, you could try to generate your antibodies with a fusion construct and then remove fusion-specific antibodies by affinity binding to GST/MBP alone. It would be labor-intensive, though, requiring lots of binding steps, and you may not be able to fully purify it away.
The other option is to make the fusion construct cleaveable and see if you can generate soluble protein after partial purification and removal of the fusion. It's tricky, but it can work in some circumstances.
Thanks! I figured i'll try to see if i can purify the partially soluble HIS-tagged protein anyway, before trying other fusion contructs or truncations. But the protein just came out of the column as soon as I put it in, indicating it wouldn''t bind to the Ni-NTA column. I think i know what the problem is (didn't set the beads in 8M urea buffer and didn't wash or elute the protein in 8M urea buffers...so the protein probably pelleted back)....my question now is if I need to include Imidazole in the 8M urea containing buffer (cos the QIAexpressionist protocol does not say to include...their protocol only includes a gradual reduction in pH)..how does this result in elution of protein from the Ni-NTA beads?
Hi ggUss,
I want to know your protein contain any disulfide bond? Because it is hard to form disulfide bond into E.coli's cytoplasmic and result the inclusion bodies.
Second, I want to know when you express the protein at 16 degree, did you shake (strong or weak)? If you shaked strong, you could low the rate (even no shake) and express overnight.
And I think you should use Guanidine-HCl 6M to make a soluble of inclusion bodies (it is better for protein from Eukaryote).
Good luck! 
I want to know your protein contain any disulfide bond? Because it is hard to form disulfide bond into E.coli's cytoplasmic and result the inclusion bodies.
Second, I want to know when you express the protein at 16 degree, did you shake (strong or weak)? If you shaked strong, you could low the rate (even no shake) and express overnight.
And I think you should use Guanidine-HCl 6M to make a soluble of inclusion bodies (it is better for protein from Eukaryote).
Good luck!
Hi chukynam,
I have no clue as to whether my protein contains a disulphide bond (signs of a bad scientist???!!!) but would be grateful if you can tell me if there's a website I can go to to find out.
When expressing @ 16 degrees, I shook at 230 rpms...pretty fast...so yes, I will try expressing overnight at maybe about 50rpms or so? (without shaking the bac will settle right?)
I've also heard about 6M Guanidine -HCl, but a protocol that i read included an expensive amino acid (arginine?), which led me to try the less expensive 8M urea first. Can you please refer me to a good, tried and tested protocol? I would truly appreciate that. Thank you for your insight.
Regards,
ggUss
pH affects the protonation state of the histidine. Above pH 8, the histidines are mostly stripped of protons, and therefore have free electron pairs to donate to the metal. Below pH 8.0, they're become partially protonated. Their free electrons get wrapped up in binding a hydrogen, so they'll just pass right by the metal.
If you follow this protocol, you shouldn't need imidazole except to remove weaker binding contaminants in the preliminary wash. Either one will elute your protein (imidazole by direct competition, pH by changing the protonated state of His).
pH affects the protonation state of the histidine. Above pH 8, the histidines are mostly stripped of protons, and therefore have free electron pairs to donate to the metal. Below pH 8.0, they're become partially protonated. Their free electrons get wrapped up in binding a hydrogen, so they'll just pass right by the metal.
If you follow this protocol, you shouldn't need imidazole except to remove weaker binding contaminants in the preliminary wash. Either one will elute your protein (imidazole by direct competition, pH by changing the protonated state of His).
Thanks aludlam! Its always nice to know the theory behind a lot of the lab stuff that has been tried and tested and that i tend to take for granted. I tried the reduction in pH, which eluted the protein well. But there was still some non-specifc protein binding in the elution fraction containing the majority of my protein. So, I might try including imidazole at a low concentration to try and remove these.