Hi! I'm trying to express a human protein (GST-NTCP) in E.coli (BL21 cells). The protein is cloned in pGEX6P-1 vector and is 64KDa in size. After DNA cloning, the plasmid's sequence was correct so we proceed to regrow the bacteria in 5ml of LB/amp buffer overnight. Then, we measured the absorbance using Pierce BCA kit and ran a SDS-PAGE gel using GST as positive control. On the western blot, GST band was the right size and very bright but, the GST bands on the other clones were smaller. On the 1 min exposure I have no other consistent band and on the 2-3 min exposure I have a lot of background. (See attached pictures)
How can I overcome this problem. I don't even know what I'm doing wrong. Thanks so much! :)
Hello Alana, a few things are not very clear in your description (for example are you inducing at all? are you inducing O/N? What type of lysis are you doing... P and S are pellet and super, so possibly a native lysis?). Anyway, my suggestion is to proceed step by step, first try to get some expression, second verify solubility. So, I'll try to put it in order:
A. Expression:
Expression is influenced by the bacterial strain, the temperature, the time, the antibiotic, the medium and the amount of inducing reagent (IPTG with your pGex6P vector).
1) You are using BL21 cells, they have no aiding plasmids expressing tRNAs for difficult eukaryotic codons, so depending on the sequence of your human protein (rich in lysine, arginine, proline, etc.) you might never have enough expression... Just to keep in mind in case you will not succeed (then you will have to try different bacterial strains optimized for eukaryotic expression).
2) I assume you are using standard LB. Fine, just keep it in mind, expressing proteins requires a lot of "food" and LB is a very basic medium. If you will have a moderate expression a way to push it up will be using richer media (there are many, but already the 2Y that doubles the amount of yeast extract will help...).
3) The antibiotic you are using is ampicillin, which is fine in principle, except that when you induce protein expression you do something that bacteria don't need and don't like, so their reaction is to get rid of the plasmid. With ampicillin this is very easy, because this beta-lactamic antibiotic is very unstable and temperature sensitive, so after a few hrs at 37 C the amount of ampicillin in the medium is so decreased that bacteria can do without a beta-lactamase and will get rid of the plasmid, so no expression. For expression it is strongly suggested to use carbenycillin, 100-200 ug/ml, a much more stable antibiotic that will ensure bacteria keep the plasmid till the end.
4) IPTG concentration can have an effect on induction of protein expression. Typically you could start with 1 mM, but you could try at least 0.5 mM and 2 mM if things don't work at all or work too little with 1 mM.
5) Now the strongest factors, let's start with time. This is fundamental, every protein has its own optimal expression time. To determine it you have to run a time course of expression. How do you do it. Not too difficult. Inoculate one (better 2-3) BL21 clones from an agar plate into 2 ml of LB-Amp each and culture O/N at 37 C(without induction, just to grow the cultures). In the morning aliquot 20 ul of this O/N culture into 6 tubes with 2 ml of LB-Carb (1:100 dilution) and grow them 2-3 hrs at 37 C until they have OD 6-800 (you can prepare an extra tube for the OD measurement, typically 2 ml will allow you to check twice the OD of your culture). At this point set aside one aliquot for lysis as non-induced sample (we'll see the lysis in a moment), and induce the other 5 aliquots adding 1 mM IPTG and culturing for 2, 4, 6, 8 hrs, and O/N at 37 C, setting aside one aliquot at each time point for lysis.
How to lyse? In a first approach do a total lysate, if you will get expression you'll work to see if the protein is soluble or not. This means simply cooling down on ice the 2 ml culture for 10 min, transfer in 1.5 ml tube, spin down quickly to pellet bacteria, remove LB and add 200 ul of 2x Laemmli buffer. Pipette to lyse and boil 10 min at 99 C vortexing briefly every 2-3 min to shear DNA if necessary (you'll see it when you pipette, if the solution is very viscous and filamentous you have a lot of cells and the gDNA makes this effect. Vortexing will break the big DNA molecules into smaller fragments and the lysate will become fluid).
Once you have all the lysates, you can quantify if you wish, or load 20 ul on gel, uninduced and the 5 induction points all next to each other (possibly 2-3 clones, as mentioned above, one gel per clone), with your protein MW marker. Then just stain the gel with Coomassie Blue stain and look if something appears at the height of your fusion protein in any of the induced samples.
6) As said, temperature can be very important. If no expression is obtained at 37 C with the procedure above, try the induction protocol (so from when you add IPTG mid morning of the second day) at room temperature and 15 C (sometimes even 30 C is already very effective).
Keep in mind that you may have leakage from the promoter, this is not uncommon. This means even uninduced cells might express some of your protein, so by Coomassie you might see already some prominent band at the height of your fusion protein. There is no absolute control on expression, but you should experience that some clones leak more, some less.
Don't be surprised if there is no WBing involved. If you need WB to see tour protein it means the level of expression is definitely too low. Coomassie stain must be enough to see the extra band growing in intensity in the induced samples.
Finally if induction works, the bacteria stop growing, so the OD should remain constant. This is because they are devolving all their energy and resources into producing your protein. This also means you should not incur into overloading the gel, normally from the first to the last sample the proteins on the gel are pretty much the same. If this is not happening and the proteins increase with the induction time this is a sign that induction is not working, so check IPTG, check temperature, eventually check the bacterial strain.
B Solubility.
This is a completely different problem. Often solubility of eukaryotic proteins in bacteria is impaired. Again, there are strains optimized for this, which express chaperones. Anyway first test it on your protein.
Testing solubility is not difficult, you were already doing it. Choose the induction point/protocol that best expresses your protein to start with, and do a native lysate (with triton or similar detergent and sonication to shear the DNA... but be very careful, sonication easily overheats the lysate and denatures proteins (which will then aggregate and pellet), so you have to keep your samples on ice and do very short bouts of 5-10 s max interspaced by 30-60 s at middle power), spin the lysates 10 min at 4 C and collect super and pellet. Again it should be sufficient to laod a gel and stain with Coomassie. My suggestion is to have in parallel an uninduced and induced culture for a total lysate to load next to the native one to verify the induction.
If your protein is not soluble, you can try a few things. You can step back to earlier induction points if possible: often if you have less protein it folds better, when the production is greater it aggregates more easily before it can fold, just a time/concentration effect. Other possibility is to lower the culture temperature, typically this gives more time to the protein to fold and by reducing kinetic energy favours intramolecular interactions (folding) towards intermolecular ones (aggregation=precipitation). Eventually you might need to change bacterial strain or move up to eukaryotic expression systems.
I hope this will help. Good luck.
Hello Alana, a few things are not very clear in your description (for example are you inducing at all? are you inducing O/N? What type of lysis are you doing... P and S are pellet and super, so possibly a native lysis?). Anyway, my suggestion is to proceed step by step, first try to get some expression, second verify solubility. So, I'll try to put it in order:
A. Expression:
Expression is influenced by the bacterial strain, the temperature, the time, the antibiotic, the medium and the amount of inducing reagent (IPTG with your pGex6P vector).
1) You are using BL21 cells, they have no aiding plasmids expressing tRNAs for difficult eukaryotic codons, so depending on the sequence of your human protein (rich in lysine, arginine, proline, etc.) you might never have enough expression... Just to keep in mind in case you will not succeed (then you will have to try different bacterial strains optimized for eukaryotic expression).
2) I assume you are using standard LB. Fine, just keep it in mind, expressing proteins requires a lot of "food" and LB is a very basic medium. If you will have a moderate expression a way to push it up will be using richer media (there are many, but already the 2Y that doubles the amount of yeast extract will help...).
3) The antibiotic you are using is ampicillin, which is fine in principle, except that when you induce protein expression you do something that bacteria don't need and don't like, so their reaction is to get rid of the plasmid. With ampicillin this is very easy, because this beta-lactamic antibiotic is very unstable and temperature sensitive, so after a few hrs at 37 C the amount of ampicillin in the medium is so decreased that bacteria can do without a beta-lactamase and will get rid of the plasmid, so no expression. For expression it is strongly suggested to use carbenycillin, 100-200 ug/ml, a much more stable antibiotic that will ensure bacteria keep the plasmid till the end.
4) IPTG concentration can have an effect on induction of protein expression. Typically you could start with 1 mM, but you could try at least 0.5 mM and 2 mM if things don't work at all or work too little with 1 mM.
5) Now the strongest factors, let's start with time. This is fundamental, every protein has its own optimal expression time. To determine it you have to run a time course of expression. How do you do it. Not too difficult. Inoculate one (better 2-3) BL21 clones from an agar plate into 2 ml of LB-Amp each and culture O/N at 37 C(without induction, just to grow the cultures). In the morning aliquot 20 ul of this O/N culture into 6 tubes with 2 ml of LB-Carb (1:100 dilution) and grow them 2-3 hrs at 37 C until they have OD 6-800 (you can prepare an extra tube for the OD measurement, typically 2 ml will allow you to check twice the OD of your culture). At this point set aside one aliquot for lysis as non-induced sample (we'll see the lysis in a moment), and induce the other 5 aliquots adding 1 mM IPTG and culturing for 2, 4, 6, 8 hrs, and O/N at 37 C, setting aside one aliquot at each time point for lysis.
How to lyse? In a first approach do a total lysate, if you will get expression you'll work to see if the protein is soluble or not. This means simply cooling down on ice the 2 ml culture for 10 min, transfer in 1.5 ml tube, spin down quickly to pellet bacteria, remove LB and add 200 ul of 2x Laemmli buffer. Pipette to lyse and boil 10 min at 99 C vortexing briefly every 2-3 min to shear DNA if necessary (you'll see it when you pipette, if the solution is very viscous and filamentous you have a lot of cells and the gDNA makes this effect. Vortexing will break the big DNA molecules into smaller fragments and the lysate will become fluid).
Once you have all the lysates, you can quantify if you wish, or load 20 ul on gel, uninduced and the 5 induction points all next to each other (possibly 2-3 clones, as mentioned above, one gel per clone), with your protein MW marker. Then just stain the gel with Coomassie Blue stain and look if something appears at the height of your fusion protein in any of the induced samples.
6) As said, temperature can be very important. If no expression is obtained at 37 C with the procedure above, try the induction protocol (so from when you add IPTG mid morning of the second day) at room temperature and 15 C (sometimes even 30 C is already very effective).
Keep in mind that you may have leakage from the promoter, this is not uncommon. This means even uninduced cells might express some of your protein, so by Coomassie you might see already some prominent band at the height of your fusion protein. There is no absolute control on expression, but you should experience that some clones leak more, some less.
Don't be surprised if there is no WBing involved. If you need WB to see tour protein it means the level of expression is definitely too low. Coomassie stain must be enough to see the extra band growing in intensity in the induced samples.
Finally if induction works, the bacteria stop growing, so the OD should remain constant. This is because they are devolving all their energy and resources into producing your protein. This also means you should not incur into overloading the gel, normally from the first to the last sample the proteins on the gel are pretty much the same. If this is not happening and the proteins increase with the induction time this is a sign that induction is not working, so check IPTG, check temperature, eventually check the bacterial strain.
B Solubility.
This is a completely different problem. Often solubility of eukaryotic proteins in bacteria is impaired. Again, there are strains optimized for this, which express chaperones. Anyway first test it on your protein.
Testing solubility is not difficult, you were already doing it. Choose the induction point/protocol that best expresses your protein to start with, and do a native lysate (with triton or similar detergent and sonication to shear the DNA... but be very careful, sonication easily overheats the lysate and denatures proteins (which will then aggregate and pellet), so you have to keep your samples on ice and do very short bouts of 5-10 s max interspaced by 30-60 s at middle power), spin the lysates 10 min at 4 C and collect super and pellet. Again it should be sufficient to laod a gel and stain with Coomassie. My suggestion is to have in parallel an uninduced and induced culture for a total lysate to load next to the native one to verify the induction.
If your protein is not soluble, you can try a few things. You can step back to earlier induction points if possible: often if you have less protein it folds better, when the production is greater it aggregates more easily before it can fold, just a time/concentration effect. Other possibility is to lower the culture temperature, typically this gives more time to the protein to fold and by reducing kinetic energy favours intramolecular interactions (folding) towards intermolecular ones (aggregation=precipitation). Eventually you might need to change bacterial strain or move up to eukaryotic expression systems.
I hope this will help. Good luck.
NTCP is Na+-taurocholate cotransporting polypeptide?
If so, then expressing membrane proteins in E. coli can be tricky since the pathways for membrane protein maturation are significantly different in prokaryotes compared to eukaryotes. Has this protein be expressed in E.coli before? If not then you might think about changing expression systems to Pichia, insect cells or another eukaryotic cell line.
Good luck
- Badges
- Science method