Negative Controls for KO Cell Lines

[Joey]

I am currently using pSpCas9(BB)-2A-GFP (PX458) to make knockout cell lines. I’m not entirely clear on what the best negative control would be. I was thinking of simply using the parental cell line as a negative control for downstream functional assays. 

Another option is to express the empty vector in my cell line to control for Cas9 overexpression. However, from my understanding, these empty vectors would still produce a functional guide sequence because of the DNA directly in between the U6 promoter and sgRNA scaffold (containing the dual BbsI sites). Therefore, this sgRNA that is expressed for the U6 promoter of an empty vector could theoretically target specific genomic loci and cause off-target effects.

I could also transfect my cells with a PX165 (which only expresses Cas9 without a sgRNA). This seems like it would be more robust control for Cas9 overexpression. Alternatively, I was thinking of cutting the empty vector  (PX458) with BbsI to remove the sequence between the U6 promoter and sgRNA scaffold and express this in cells for a negative control.

If anyone has any advice on this topic it would be greatly appreciated.

Thanks,

Joey

[Benjamin Gowen]

Hi Joey,

I think the best negative control would be a vector expressing a functional guide RNA that doesn't target any sequences in your target genome. That way you don't have to worry about "specific" cuts from the guide RNA produced by the empty vector. I'm not sure what the literature currently has to say about the behavior of Cas9 in the absence of guide RNA expression, but having a guide present seems like a better control than no guide at all. It's also possible that expressing tons of guide RNA could have Cas9-independent effects (maybe it dominates the processing machinery, or the scaffold region binds some other nuclear protein). To be clear, I'm not aware of any evidence that these are serious concerns, but if you're looking for the best negative control, I think expressing a non-targeting guide RNA is the way to go.

Benjamin

[Joey]

I think expressing a negative control sgRNA (targeting GFP for example) is the best way. I've seen multiple studies that use this approached.

Thanks for the advice Benjamin, much appreciated :)

Joey

[Chris Thorne]

I think using the matched parental downstream is the way to go. As you're only transiently transfecting the plasmids, and deriving a clonal cell line can take 8-16 weeks, I'm not sure making a non-targeted clone independently adds a great deal of value -  it is not the presence of the Cas9 that needs to be controlled, but the changes in the genome. Unlike RNAi where you are controlling for the off target effects of transfection or expression of shRNA, for genome engineering the original cell line from which the engineered clone is derived serves as matched isogenic, non-engineered control for the line. 

It's worth bearing in mind that a proportion of your clones will not be targeted, and that being the case may in fact serve as the best control should you wish to use them - as they will be subject to the same offtarget potential

Chris

[Joey]

Hi Chris, 

I really like that idea of using non-targetted clones as negative controls. I guess if I'm going to do it this way it's best to generate a non-targetted/WT clone for each sgRNA I use. So in that case I'll have multiple, independent non-targetted/WT clones to use as negative controls. I think this will serve as a robust control for my downstream functional assays. 

Thanks for the suggestion. 

Joey