I'm new to siRNA and so have a number of questions, but I'll start here:

1. Reading the postings on this part of the forum it would appear as though scientists are comparing different transfection reagents and using knockdown of either GAPDH etc or their target gene expression as a readout. Don't you optimize the transfection using fluorescently labelled siRNAs?

2. People quote transfection efficiencies of say 60% and then say that this gives them >80% knock down. This suggests that; (i) In transfected cells, 100% of the target gene's expression is inhibited, and (ii) expression of the target gene is also inhibited in cells that were not transfected. Is this the case?....or,

Is this because of a reduced viability of some of the non-transfected cells or because the transfection efficiency is being under estimated ( ie. whereas it might be possible for 10 molecules of duplex siRNA to inhibit some expression, it would not be possible to detect these if they were fluorescently labelled)?

Thanks for any answers to these questions.

You're really compare apple with orange. 60% efficiency means 60 cells out of 100 cells uptake siRNA (which covers all kinds of cells getting a few, some to many siRNA....). The 80% knockdown means the inhibitory effect on "steady-state" RNA or protein level, and you can expect the half-life, the sampling time, the target property, cell-cell talk (and many other things...) would affect the final readout. So, the knockdown effect doesn't necessarily be less than tansfection efficiency!

Thanks for your reply but it still doesn't make sense:

If the transfection efficiency is 60% then clearly 60 out of 100 cells have taken up at least the minimum amount of siRNA. However, to get an >80% reduction in steady-state RNA levels you must be affecting MORE than just the transfected cells. And, you must be completely knocking down expression in all of the transfected cells.

It sounds like there must be some additional effect such as toxicity of the transfection reagents that kills some untransfected cells and/or the actual transfection efficiency is being grossly under-estimated. I think both may be true.

I would like to know how people arrive at transfection efficiency readouts. Do they use flourescently labeled siRNAs? If so how do they differentiate between cells that have bound the siRNA and those that have actually internalized it? (confocal microscopy?)

And, how many FITC-siRNA duplexes do you have to get into a cell before you can detect a signal? I asked a technical guy from NEB (who sell the FITC-siRNA) but he never got back to me.

I agree that it is strange, and I agree it is apples and oranges.
Transfection efficiency is usually judged by eye (fluoresecent siRNA, loss for b-gal staining). But transfection with minute quantities of siRNA may still have effects (but the cell appears not transfected under fluorescence microscope) and silencing of 20-40% of b-gal is not detected (still stains quite blue). When you quantify, you usually use something like RT-PCR, Western blot, FACS, then these effects are taken into account. That's my explanation at least.

I think the conclusion is that fluorescently tagged siRNAs are no good for determining transfection efficiency. The only quantitative readouts are RT-PCR and Western blot, although FACS has the added advantage of being able to quantitate expression level for each cell. I have the feeling that it should be possible to get some idea of the transfection efficiency by looking at the spread of fluorescence intensities following FACS. There should be a normal distribution for the number of siRNA complexes taken up by the cells and some form of reverse extrapolation should tell you what percentage of the cells are likely to have had no siRNA uptake.
In the end, transfection efficiency is not the critical factor and buying NEB's FITC-siRNA was a waste of money. I think the reason the NEB rep didn't answer my enquiry was that he'd heard it all before.

Guys,

Another possibilities: IFN or off-target effect which may affect the expression of your internal control (GAPDH, actin..), and lead to a more profound silencing. so, it is possible that 60% efficiency result in >80% inhibition.

Off-target effects could be possible if the off-target gene is linked to your target gene (whose expression you are monitoring). How likely this is would depend on the target gene.

Usually siRNA is chosen to be 21-23 nt long specifically to avoid an IFN effect (which occurs with longer dsRNAs)

My favourite explanation is still that the detection limit for transfection efficiency estimation is much higher than the knock-down activity. ie. 5 siRNA duplexes will knock down expression to some degree, but 5 FITC-labelled duplexes would not be visible using a fluorescence microscope.

I agree that the difference in detection limits for estimating transfection efficiency and determining expression levels may be responsible.
Another consideration is that different siRNAs may differ in their transfection efficiency. Are people using FITC-labeled siRNAs against their target, or just some irrelevant FITC-labelled oligo to estimate transfection efficiency?