NgAgo for genome editing

Hello Everybody,

Maybe some of you have heard about this potential new way to edit genome using the protein named NgAgo.
(More details here http://www.nature.com/nbt/journal/v34/n7/full/nbt.3547.html)

This protein was looking pretty great, because it uses 5’P-ssDNA as guide (no cloning or transcription needed), no PAM needed (any place in the genome could be targeted) and seems to have way less off-target activity than the Cas9 most of us are currently using.

…Unfortunately, this discovery, which could have been big in the genome editing field, seems to be problematic and many labs have hard time to reproduce the published data.
(More details here http://www.ipscell.com/2016/07/gaetan-burgio-new-data-theory-on-ngago-versus-crispr/ and https://pubpeer.com/publications/27136078)

So, just to be transparent and maybe avoid people to waste their time, we tried it in the Cochella lab.

  • We ordered a codon optimized version of the NgAgo enzyme.
  • We then have made a eft-3p::ngAgo(opt)::tbb-2 3’UTR construct (taking as a template plasmid developed by the Calarco lab (https://www.addgene.org/46168/)). Strangely this step was already really problematic for some reasons.
  • We then tried to mutate the dpy-10 or unc-119 coding sequence using 5’P-ssDNA which were targeting the same place as previously published for the CRISPR approach (so these sites should be accessible for editing and should work pretty fine).

And to cut a long story short, we tried several time and so far we never got any mutant worms.
To be honest, we did not try super hard, but with the numbers of injected P0 and F1 we got, if the method was working, we should have found couple of mutants.

So we were wondering if other people have tried it as well, and if they obtained better results, or if they face the same issues ?

Also, if nobody as tried it yet, I would strongly suggest to wait that the problem linked to the original paper are fixed and see what are the requirement to have ngAgo working for genome editing.

I hope this helps.

Cheers.

Many thanks for sharing your experiences with the community. It’s extremely helpful to have people like you and Gaetan Burgio share your efforts. I had considered trying NgAgo, but was a bit concerned that the oligo had to be loaded as soon as the protein was translated, so wondered if there might be a limited window to create an editing competent NgAgo in the germline. I’d asked the group that published the method for a small sample of protein and their expression construct (as it was not deposited in AddGene), but no luck (had been wondering if a protein-oligo complex akin to Cas9 RNPs would be best). It really seems that NgAgo is either not as good as advertised, or very finicky. It’s really awesome to have people sharing negative data on the forum!

Thanks Jordan for the comment. I do believe that the only way to go really further is to do open science, but it seems that not everybody likes it :wink:

Anyway, to be honest, we did not push the project too far (not as far as Gaetan Burgio), we could have go a bit further and try couple of different things (different plasmid concentration, a lot of different primer concentration…), but after seeing the mess done by NgAgo in the scientific community we decided to step down, share with everybody our current observations and see what is coming .
That is also why, I am asking here if anybody tried it in the worm more seriously than we did, to be sure that we did not simple messed up…

And just to convince us that we did not messed up that bad, I am pretty sure that labs like the lab of Feng Zang would have already communicate something about it if this would be working + number of lab have been unsuccessful with NgAgo so it is pretty clear that there is a problem somewhere.
Also, at the same time we tried it in the worm, the facility next to our institute was currently trying it for mammalian cell and so on, using purified protein and they also told us that they are getting troubles…
So these are many elements that are showing us that indeed NgAgo has been a bit overhyped…

To come back to your come back to your comment about the oligo loading. We considered this issue, from the original paper and discussion with the facility mentioned above, it seems that the loading is extremely fast and stable.
So if we provide excess of oligo in the germline it should be fine, then the limited factor is the transcription efficiency of the eft-3::NgAgo in the germline, which could be hard to control.
An other issue we thought about is the speed of the reaction, it is not clear if the cleavage can be done fastly, so this could be one the biggest issue to make it working, if the cleavage takes a while even though the complexe is formed, it will be pretty hard to get cutting events.

But all this is purely speculative, as many of us we will stick to CRISPR which works really well, just in my case with duplicated genes finding sgRNA that targets precisely only one is a challenging task…NgAgo would have solve many of the issues I can face, that was also one of the reason why we jumped on the NgAgo hype-train :wink:

Cheers.

A pretty interesting survey about the NgAgo Fiasco
https://medium.com/@GenomicsEagles/the-global-experimentation-on-ngago-4f0e282c2190#.hm1cylgmx
Cheers.