Hi Elanor,

cloning 4 gRNA with 4 promoters should technically work, but a more reliable approch would be "Pol II promoter + tRNA processing" —your transcript includes cleavable elements like tRNAs or Csy4 that are processed into individual gRNAs. the design would require some research but the transcription should be more robust.

see this --> DOI: 10.1007/978-1-4939-8991-1_5

That's a rather big plasmid and cloning could be a bit challenging but insertion of DNMT3a also is very doable. Good Luck

Hi Elanor,

This can be done in several ways. I've cloned up to 4 working gRNA under the control of a U6 promoter using tRNA spacers for transcript processing.

Article (Po)STAC (Polycistronic SunTAg modified CRISPR) enables live...

In this paper however we only used 2 gRNAs, but I have successfully used the method to clone up to 4 different ones.

I'm not sure you would get the same efficiency with a pol II, mostly for RNA subcellular localisation and processing.

However, while this approach is doable and it's saving some space on the vector, cloning will be quite a journey. Most DNA synthesis company won't touch this with a pole and you'd have to assemble this via overlap extent PCR from templates or some other labour intensive cloning approaches. Nothing too exotic but it can put you off and cause delays.

Given the cost of DNA synthesis however, the quickest thing is to synthesise whole U6 gRNA cassettes and link them via homology regions (20-30 bp) via Gibson assembly (or golden gate if you prefer).

This should give you the highest expression level per gRNA, although it might hamper the stability of your plasmid due to repetition of identical sequences (U6 and scaffold). Bear in mind that the scaffold will be repeated also if you choose a multiplexed approach with a unique promoter.

As long as your plasmid is not a lentiviral backbone you shouldn't have stability problems in most cloning strains (DH5a or Top10). I have cloned up to 8 U6 promoters and gRNAs on the same plasmid and it works. Saves you time on the design and it's easier to do it with synthetic fragments.

If you use TwistBioscience each cassette should cost you around 30$, which is affordable and fast.

For DNMT31 insertion digest with NheI NotI and excise the Tet fragment, you can replace this with a synthetic fragment encoding your DNMT31. Make sure it's in frame with Cas9 ORF and include overlapping ends (20-30 bp) with the linearised backbone. You can then use gibson to generate your construct.

It would probably be possible to clone DNMT31 alongside the 4x U6 cassettes in the same cloning round if you don't want to generate an intermediate. You'd have to change the restriction enzymes used.

Hope it helps

Francesco

Forgot to add something, if you opt for a multiplexing strategy with a single promoter you cannot have terminators spacing your guides, or you'll just end up producing the first guide.

In our multiplex these must be substituted by tRNAs, with a final terminator at the end of the cassette after the 4th gRNA.

Hope it helps

That's incredibly helpful Francesco thank you so much!

I'm considering using the Gersbach lab system (DOI:10.1093/nar/gku749) where they use hU6, mU6, H1, and 7SK promoters to reduce the risk of recombination, do you think this is a good approach to take?

When it comes to ordering the pieces, is your recommendation then to order each promoter/gRNA/scaffold/terminator construct separately with homologous overhangs for Gibson assembly? Or would the use of differing promoters perhaps make it less daunting to a DNA synthesis company?

Thanks so much for the tips about DNMT3a as well, I'll be using primers for gibson assembly that have been tried and tested by another group so I'm hoping they should work as intended.

Thanks again,

Elanor

Hi Elanor,

In Gersbach et al. I think the issue was that they were using lentiviral backbones as final destination vectors. These are extremely recombinogenic in E. coli, particularly when repeated sequences are introduced. The problem can be partially fixed by using Stbl2/Stbl3 cells grown at 30 C but the best is to avoid repeated sequences altogether.

This explains the rationale behind using U6 promoters from different species however, in my view, this only partially solves the issue. Although you can modify the promoters, you'd still have 4 identical scaffold sequences, which could undermine the stability of your vector.

Rest assured that this problem, in my experience, only applies to lentiviral backbones (and potentially to AAVs which, to some extent, are plagued by similar issues in E. coli).

If you are working on a standard cloning backbone (with no LTRs or ITRs) you shouldn't have any issue. I have cloned up to 8 U6 promoters/gRNA cassettes on the same vector, without issues in stability or plasmid propagation using standard E. coli strains. The sequence for this vector can be found in the paper below, Supplementary Table 1 plasmid 6.58.

The work is described in this paper:

https://academic.oup.com/nar/article/50/13/7783/6633882

In Supplementary Table 1 you will find all plasmid sequences. You can use the sequence of plasmid 6.45 as a template for cloning quadruplex gRNAs. You can choose wheter to use 4x U6 or opt for the Gersbach approach depending on your cloning backbone.

For the implementation you order synthetic sequences without adaptors from Twist. Each cassette will need to be designed as:

1) 5BH - U6/gRNA1/Scaffold - L2

2) L2- U6/gRNA2/Scaffold - L3

3) L3- U6/gRNA3/Scaffold -L4

4) L4- U6/gRNA4/Scaffold - 3BH

5BH - 30 bp homolgy to backbone (5' portion)

L2/L3/L4 - 30 bp linkers

3BH - 30 bp homology to backbone (3' portion)

You can use any linker you choose, or implement those used in plasmid 6.45 from my paper. The best is to include unique restriction sites in each junction so you can easily screen your colonies.

Best wishes

Francesco

Hi Francesco,

Apologies for the late response. That is amazingly helpful, and I am very grateful! I'll definitely be using that sequence, and I very much appreciate the time you put into your answer. That paper will be being cited in my work, and I'll likely use that plasmid as a template for my multiplexing!

All the best,

Elanor