Low expression: What else can I do? - Using pET23a in BL21 PlysS (Aug/23/2009 )

Hi there!

My intention here is not open another topic about low expression (more HERE). Since I haven't found the same conditions I've been using here is the problem:

Expression of 2 bovine proteins, using contructs based on pET23a(+) with His-tag. Both are been expressed in BL21 PlysS wiith leaky and low expression. When I check with SDS-PAGE I can't distinguish my protein from the rest, and the purification steps fail.

I've tryed:

Different IPTG concentrations: 0.1 0.3 0.5 1.0 mM
Different temperatures: 37°C and 30°C
Different media: 2xYT and LB
Induction times: 3, 6, 9, 18, 24 hrs.

What else should I try?

Thanks a lot!

1. I presume you have checked the DNA sequence to make sure you have absolutely no errors (frameshifts etc).
2. Do the sequencing again.
3. Have a look at Studier's autoinduction expression system, it should fix the leaky expression problem.
4. What is the induced band at ~30kDa?

4. No idea!!!! I wonder if it is a protein dimer...

We have not had much success ourselves with the pLysS strains, so you are not alone there. Since you are getting some expression, I will presume the vector is good. As I could not view your blot, I could not determine if the antibody was anti-His tag or anti-target protein. If the former, then the data would indicate that the fusion to the tag sequence is correct and the problem lies with expression. My team has a few years of experience dealing with similar problem proteins so short of using a different expression vector (which you might want to consider too) and at the risk of being redundant, here is how we have approached solving this:

First, test for expression using an in vitro expression kit (it is really simpler that most folks realize). This will help you diagnose whether the problem is with the ability to express the protein or the interaction of the protein with the host. If the protein can be expressed in vitro, then test for accumulation in different strains. We'll use up to six different strains with diverse genetic backgrounds. There are some new strains available, derived from BL21 known as C41 and C43, that were isolated as lowering the toxic effect of some proteins. (However, did you selected a host harboring pLysS based on evidence of a problem in another host?) After selecting the best strain, test different media compositions; not just 2xYT and LB. We have found that the nitrogen source as well as the carbon to nitrogen ratio can influence the level of recombinant protein accumulation so you might want to try some of the other formulations that are available (let me know if you need some ideas). If the above two analyses improve expression, then revisit the culture and induction conditions to optimize production.

If the in vitro expression does not give you good production, then the problem lies with the expression system or there is a problem sequence in the coding sequence of your protein. You might considered alternative approaches to expressing proteins in E. coli. For example, fusion to a bacterial protein often helps if this doesn’t create a problem with the scale of your downstream purification or with your intended use. The other approach is to secrete the protein to periplasm or if possible to the medium. The downside here is more work constructing a new vector.

Finally, there is the brute force approach. We have had a couple of cases where the best approach was to run a large-scale fermentation and apply the extract to a small column. However, the protein needs to stick to the affinity resin for this to work, which by your comments I suspect might be a problem. This is very rare though has happened where the tag is sequestered or sterically hindered. The use of a non-denaturing concentration of urea (<4 M) might help. If the immunoblot is showing the presence of a tag, you could try purifying under denaturing conditions to verify that the protein can be purified. The downside of then purifying under denaturing conditions, which will almost always work, is that you would have to figure out how to refold the protein.

Good luck.

Sorry taking so long to answer. Thanks for your comments, really. I'd like to try new media, what would you suggest?

I'll consider the in vitro expression kit. I never used one. Can I use my plasmid or I must construct another one?

Thanks again.

If the codon usage of your gene differs significantly from that of E. coli, perhaps one of the CodonPlus strains from Stratagene would help? I got much better expression of a plant gene and of a human gene when I used a BL21 (DE3) CodonPlus host rather than just BL21 (DE3)...

I've already tested the Codon PLus, but maybe I should try again with some clones.

For the medium alternatives you basically want to try different formulations that uses glucose or glycerol as the carbon source in combination with different protein hydrolysates. Of the latter, casein and soy derived products are the most popular. Use a phosphate buffer base such as M9 salts and build on that. Remember, that if you use 1% glucose you should increase the buffering to counter the anticipated drop in pH. A matrix design will allow you to identify which components are most critical and then optimize from there. I recommend starting with 0.4% Glucose M9 and 0.4% Glycerol M9 supplemented with different hydrolysates (use 10-20 g/L). Terrific Broth is another good selection and will often work when LB does not.

This is a lot of work and something we use to do routinely until we noticed that several formulations were consistently superior to the standard LB and Terrific Broths. At the risk of offending everyone, our company (under the Athena Enzyme Systems brand) has made these different media formulations available in a kit format to reduce the screening to one experiment. We also have some other tools to help folks improve the expression and recovery of recombinant proteins. I hope you are not uncomfortable with the commercial, but developing methods for improving the expression of recombinant proteins is what we do for a living. There are some white papers loaded with relevant references along with some protocols on our tech support pages that you might find helpful even if you don't what to purchase anything from us.

As for the in vitro kit, we use Promega's. There is a kit suitable for use with vectors that employ the T7 promoter.

Good luck.

Doc Sheldon on Sep 4 2009, 09:29 AM said:

Thanks for the tip. Maybe my next try will be TB medium.

Doc Sheldon on Sep 4 2009, 09:29 AM said:

Don't worry. This is perfectly OK, since you're trying to help.

Doc Sheldon on Sep 4 2009, 09:29 AM said:

Yesterday I was checking Invotrogen products, now I'm gonna take a lot at Promega's website. These kits seems to be a bit expensive. Here in BRazil grants for research are not so big, however since you don't need to use all the cell culture stuff it's OK. Oh, one more doubt: Does the in vitro expression kit substitutes the in vivo expression?

Once more, thank you!

Ivanov_br on Sep 4 2009, 08:47 AM said:

Regarding the in vivo versus in vitro, if it is a bacterial system, then yes the proteins will end up structurally identical. For mammalian systems there will be no post-translation modification in most cases.