shRNA with 100% marker expression but no knockdown efficiency at all - (Nov/21/2012 )

hello everyone. I found a very strange thing. I tried to knockdown 1 gene by two shRNAs (lentivirus), one of which has been used by two other lab members and was found good. The other one is also ok according to my experience. The hairpin is driven by U6 while it has a GFP marker driven by CMV.

To get much higher efficiency, I did something that they didn't do. I did cell sorting and isolated GFP-highly-expressing cells. Because this gene has been shown to be critical for cell growth, so after sorting, when I found that knockdown cells didnt grow well, I thought that the knockdown went quite well. However, after 1-week culture (to collect more cells), I did real time PCR and found the mRNA level was not affected at all, for both shRNAs. I also did FACS and confirmed the GFP expression was 100% and very high.

Here are some thoughts of mine but I dont know whether they are correct. I suspect that in the GFP-highly-expressing cells (among all the GFP positive cells, I picked those with even higher expression), the expression of shRNA might be hindered. It seems a little rediculous but when I knocked down another gene with a different system in which GFP is driven by hPGK promoter, I didnt have this strange thing. Could it be possible that in some case, CMV promoter is competitive to U6? I know knockdown a gene that is critical for cell growth might generate some unexpected results but the successful case that I just mentioned above is also knockdown of a cell growth promoting gene.

Have you ever had such problem? Can you provide some explanations? Thank you very much.

I think your problem will be that the high levels of GFP expression, while not essentially toxic to the cell, is using up much of the cells' resources hence the lower viability/lowered growth rate.

bob1 on Wed Nov 21 20:01:37 2012 said:

Thanks for the reply. Ok, I can accept the fact that knockdown of this gene might have no effect and what I have observed is only some side effect from highly overexpressed GFP. But why there is no knockdown at all. The hairpin is together with GFP and it has no expression at all. U6-shRNA uses pol III while CMV-GFP uses pol II. Except that the shRNA cassette was lost by some selection disadvantages. This is a homemade plasmid with CMV-GFP cassette inserted from another plasmid and the size is 10 kb. I suspect the "link" between these two cassettes is not strong enough so that it might break under some selection pressure. And finally shRNA+ cells lost GFP and I sorted GFP+ cell with no shRNA at all. This guess again seems to be nonsense. What do you think? After all, with another system (7 kb, commecially available), I did have got 100% strong GFP+ cells with good knockdown efficiency. Actually I am thinking of switching to this system.

By the way, do you happen to know any good shRNA system? I know Mission from Sigma is good but too expensive.

Sorry, don't really work with shRNA at all, so I don't know any good systems.

It could be that the shRNA is broken or that the expression of it in some manner interferes with GFP production (e.g. affects a protein needed for the production pathway).

I presume hPGK is human phosphoglycerate kinase - if so, then the expression from this system will be much (much) lower than anything running off a viral promoter such as CMV. If your hPGK vector used U6 as the promoter for the shRNA as well, then the levels of expression will be similar between the shRNA and GFP. In you current vector the GFP expression will be at least an order of magnitude higher than the shRNA.

Do you have a direct method to check that the shRNA is being expressed (i.e. not just looking at KD)?

bob1 on Thu Nov 22 20:02:47 2012 said:

thank you. In fact, hPGK is not so weak as you thought. I have read a paper showing that in driving GFP expression, it has comparable strength as CMV. But I agree that in my case, GFP expression unexpectedly didnt correlate with shRNA expression. Regarding how to measure shRNA expression directly, I, or many other people, have relied on the marker expression. But now I learned that they are not always correlated. Besides using the markers, I really dont know other methods. Anyway, I decided to switch to another system and see what will happen.
Just out of curiosity, CMV promoter is widely used because of its strong strength. Why is it strong? Is there a reason?

It is a viral promoter, these tend to be constituitively expressed as the virus "needs" the promoters running to replicate its DNA as quickly as possible. Other non-viral promoters tend to be weaker as they are often required to be controlled in some manner - if you get overexpression of some proteins you will have a cancer or the cell will die which isn't ideal. For viruses this neither of those matter (other than the cell dying before the virus can replicate) so they have to do their thing quickly...

First, do not trust to papers 100% always, even to papers in Nature journals.

Second, I take a quick look to all this page and I did not find what is your Cells.

Third, after trying some siRNA Transfection Reagents, I found that FBS in media is the only and the only problem in all failurs in siRNA Transfection. During siRNA Transfection, you should use just media without FBS for up to 5hr. Then you can add 1ml media with FBS to that. Also you should Wash cells once or twice with serum-free medium before adding siRNA Transfection Reagents + siRNA to cells.

I have used Lipo, Hiperfect and iNTerferin for THP-1 CELLS with only less than 25% Knockdown.
But when I used this from NEB, I have 70% Knockdown:

http://www.neb.com/n.../protocol25.asp

I think the most important part of that is (1) washing once or twice with serum-free medium before adding siRNA Transfection Reagents + siRNA to cells and (2) using media without FBS.

thanks you guys, learned a lot.