Your problem may come from :

-Signal Peptides : You mentioned using three different constructs, each with a signal peptide. Signal peptides are recognized by the secretion machinery of the cell, and if the protein is meant for periplasmic expression, it could be that these peptides are causing some degree of basal expression. Signal peptides can sometimes lead to higher basal expression if they're very strong or if they are cleaved inefficiently, allowing the protein to accumulate even without induction.

-The T7 Promoter : Even in Lemo21 cells, the T7 RNA polymerase (controlled by the lac promoter) can still produce some basal expression, albeit at lower levels compared to traditional BL21 (DE3). The expression of T7 RNA polymerase is still under the control of the lac operon, which can lead to low basal activity. Lemo21 cells are designed to reduce this through the lysY gene, but depending on your plasmid and construct, some basal transcription can still occur. The addition of L-rhamnose may fine-tune this process, likely by modulating the expression of the lysY gene more tightly.

-Plasmid Vector and Promoter Strength : The plasmid (pET28) typically contains a very strong T7 promoter, which may still produce some basal transcription even when the system is designed to reduce it. Sometimes, the strength of the T7 promoter can cause this uninduced expression, particularly if the protein is small, soluble, or if the system has some residual basal expression from the T7 RNA polymerase.

-Growth Conditions : You’ve already tested different growth temperatures, but it’s possible that the growth medium, inoculation conditions, or the specific batch of Lemo21 cells you're using could also contribute to the high basal expression. Some media components (such as rich media) might affect basal expression levels, so trying different media formulations could also help.

You can try :

-Optimize Plasmid Copy Number : Check the copy number of the plasmid you're using. High plasmid copy number can contribute to higher basal expression. You can try using a low-copy vector, which might help control basal expression levels. Plasmids with a lower copy number will typically have less transcriptional activity in the absence of induction.

-Modulate L-Rhamnose : Since you've observed that L-rhamnose can tune expression, try using different concentrations to optimize basal expression. It may be helpful to perform a titration experiment with different amounts of L-rhamnose to find the balance between reducing basal expression and ensuring you can still induce high expression when necessary.

-Adjust the Inducer Concentration : Even though IPTG only slightly increases expression, varying the concentration or timing of induction could help fine-tune the expression level. Sometimes, a lower concentration or delayed induction can still result in a more controlled protein expression.

-Codon Optimization : Sometimes, the issue could stem from translation inefficiencies. If your protein or signal peptide contains rare codons, it could lead to higher basal expression, especially if the system is not perfectly optimized. Running a codon optimization on your gene could reduce unnecessary translation initiation.

-Use Alternative Strains : If Lemo21 cells are not giving you the desired control over basal expression, you could try other strains with more stringent control systems, such as BL21(DE3) with a tighter repressor or strains with even stronger repression mechanisms like BL21-Gold (DE3) or Rosetta cells that help reduce basal expression.

Dear Nicolas,

Thank you very much for your clear and specific answer. I was wondering if, in this context, the use of IPTG is worthless...Since basal expression is very high and I can still use L-Rhamnose to tune it, would producing my protein without inducing it with IPTG be a problem? This should be a kind of autoinduction protocol. Do you agree?

Is there a particular reason you're using Lemo21 in the first place? This strain is used to avoid this problem specifically--leaky expression, especially if your recombinant protein is toxic. If your protein is not toxic and you're getting a good yield, there is no problem really but also no need for this particular strain.

Speaking more generally, T7 promotor is just leaky, but what you're describing (no difference between induction vs no induction) is not normal and is typically caused by a mutation disrupting the normal functioning of the lac operon, so a mutation in either lacI or lacO. If you want to further troubleshoot it, I'd do a whole plasmid sequencing to verify the presence of mutation and just reclone it to a fresh backbone (making sure it's not mutated as well).

Hi Katarzyna,

thank you for your answer. My protein is secreted and should be isolated from the periplasm to be sure it is mature and hopefully correctly folded. Lemo system is particularly suitable for secreted and membrane proteins, and for toxic proteins as well. The possibility to finely tune the protein expression by L-rhamnose is described to increase the amount of mature secreted proteins. I did not check for any LacI/Lac O sequence mutations. It's a very good suggestion. My constructs were generated by a company, and I thought that it guaranteed the correct sequence. But probably I was wrong.

Hi Elisa,

I see, in your case then the benefit of Lemo strain is the possibility of downregulation of expression. You don't want cells to overexpress because the recombinant protein will be misfolded, possibly form inclusion bodies. I suppose then it's completely fine if lac operon doesn't function properly, as long as lysY works fine. Still it's a good habit to sequence whatever plasmid you're working with, it can save a lot of troubleshooting down the road.

Hi Katarzyna,

I usually check the sequence of the cloned gene but do not consider that of the plasmid... you raised a good point.

Thank you

Elisa