I have found that tagging a protein with an NLS decreases the expression level by a factor of 2 or more. When comparing tagged and untagged constructs in HEK293 the NLS shows lower protein levels by western, and NLS-DsRed shows lower fluorescent signal than DsRed by flow cytometry. In these experiments an NES tag does not decrease expression level.
Has anyone else seen this effect?
Dear James
Please visit http://www.ncbi.nlm.nih.gov/pmc/articles/PMC155121/
Please see the abstract
The effects of inserting reported nuclear localization signals (NLSs) into the Moloney murine leukemia virus (Mo-MuLV) integrase (IN) protein, within a replication-competent viral construct, were studied. In contrast to the virus harboring IN fused to the simian virus 40 (SV40) large T antigen NLS (SV40 NLS) (J. A. Seamon, M. Adams, S. Sengupta, and M. J. Roth, Virology 274:412-419, 2000), a codon-modified SV40 NLS was stably expressed during viral propagation. Incorporation of the codon-modified SV40 NLS into IN, however, altered the packaging of the Gag-Pol precursor in the virus; viral particles contained decreased levels of reverse transcriptase (RT) and IN. In addition, the virus showed delayed kinetics of viral DNA synthesis upon infection. A panel of infectious MuLVs containing alternative IN-NLS fusions was generated and assayed for cell cycle-independent infection. Viral infection with the NLS-tagged proteins, however, remained dependent on passage of the cells through mitosis. This finding has direct implications for engineering murine-based retroviral vectors for gene therapy.
A possibility exists that NLS tagging in some cases can result in lesser accumulation of proteins due to the dependency on cell cycle
Hi James,
In my studies (which are not yet published) The basic answer to you question is no, I have not really noticed a difference at the expression level, if anything my expression is better that the GFP-protein without the NLS. At a basic level it is translated, traffics and does not accumulate anywhere weird.
The way I designed the construct was to have a triplicate NLS sequence on the N-terminal region of the protein followed by a short linker (SL) sequence to a sequence for GFP tagged protein that was already published and expression known.
The reason I probably get better expression than my non-NLS is that I designed it such that I included an enhance Kozak sequence prior to the NLS (not included in my non-nls). I moved the start codon to before this region and removed start and stop codons from the protein that was tagged before inserting to the GFP vector so mine reads:
Start-Kozak-NLS-NLS-NLS-SL-Protein-SL-GFP-stop.
Obviously expression will be reduced compared to empty GFP plasmids, this is always the case, as your vector is larger and has more specific properties e.g. translation targeting, degradation etc.
I guess there are many factors that could play into protein loss in your case including where you put the NLS and what it interferes with, similar with the GFP. In my case the GFP effect was known and the n-terminus is a short unstructured region which may have effects on protein cell function but I have not assessed as yet. So design and placement could be an issue?
Hope this helps,
S
Never heard of this. It could be a target of a nuclear localized enzyme (kinase, ubiquitin ligase, etc) that tags it for degradation. Check mRNA expression to make sure the regulation is at the protein level.
Yes, we observed same result. Do you have an answer now? Eagerly to know
- Badges
- Science method