Dear Huawei Huang

it seems that your protein contain some regions that are susceptible to proteases and it could be degraded in the E.coli citoplasm during the production.

Solution is not simple. On the basis of my past experience I think that you can try to:

1) Test different E.coli strains with low proteases; (e.g scarab express)

2) Perform N-terminal sequence of the 2 bands from SDS-page to identify the digestion site and try to stabilize it using mutagenesis

3) Use a different expression sistem (eg. brevibacillus, expi293) which may not produce the same proteases.

good luck

Manuele

Thanks Manuele! I just find it weird because no full-length protein was detected from my time-dependent protein induction result. I will try using some other hosts and see what happens. Thanks again!

Have you checked if protein might be toxic for the host? This could explain complete absence of FL.

Anyway, since you're running a lysate, you can't be sure the cleavage happens in vivo. I'd run cells on the gel, non-induced and induced, to really make sure it's not a purification issue, as PSMF is really not super effective. I'd recommend inhibitor coctails instead (if it's a purification issue).

I'd make sure you actually have an expression system issue before switching to other expression systems which can be time consuming and costly.

Katarzyna Radziwon

Thanks Katarzyna. Do you mean loading the cell directly (lysis by heating up) onto the SDS-PAGE?

You grow the cells to induction OD (typically 0.6). Then you collect 100-200 ul of the culture, spin it, discard media, and resuspend cells in a 50 ul loading buffer. Some people do it with 1X buffer, I personally don't dilute mine because excess SDS really helps with lysis. Then you heat it for maybe 15 min at ~95C. It's still gonna be very viscous, some people sonicate to chop DNA, but I just take 10 ul of the less viscous part (there will be sort of viscosity gradient there ;)) -- that's your non-induced sample. In the meantime, you induce the cells and let the expression happen. You repeat the steps with the cells right before harvesting -- that's your induced sample. Run side by side on a gel, if there is overexpression happening after induction you'll definitely see it.

If you wonder why can't your lysate be your induced sample -- hot lysis ensures all proteolytic activity is killed, so to check what actually happens in the cells you preferentially need lysate obtained through hot lysis.

Forgot to add, that this can easily be scaled down, i.e. you don't need to re-express your protein in 1L scale, you can do this in maybe 20 mL in a 250 mL flask (I wouldn't do less though to ensure comparable aeration).

Katarzyna Radziwon Thanks a lot for such a detailed answer. Definitely I would give it a try.

I’ve done both plasmid sequencing and protein identification. The sequencing result is correct, but the protein ID shows that the two overexpressed proteins match a completely different protein from Brevibacillus.

I’m really confused because this is very unexpected—we’ve never encountered anything like this before.

We also ran a SDS-PAGE using the cell before sonication. And did not see any our desired protein.

Any suggestions or advice would be greatly appreciated!

If you (over)express something from completely different strain, I'd start with troubleshooting your expression strain. There's clearly something off here. Is it possibly already transformed with other plasmid? Did you use De3 strain?

I am now afraid my strain was contaminated due to the slow growth rate of the cell containing the large (and maybe toxic) plasmid. Bacillus could be a very common contamination in the lab, which can explain why there is protein from Bacillus expressed.

Anyway, I am grateful for advice from all of you. I will redo all the things very carefully and see what happens.