A DNA-Guided Gene Editing System

In Nature Biotechnology Gao et al. (2016) describe a new gene editing system unrelated to Cas9 and based on a DNA-guided Argonaute protein with some interesting characteristics and the potential to become another useful tool.

Cas9 with functional domains highlighted in relation to target DNA and guideRNA

Cas9 with functional domains highlighted in relation to target DNA and guideRNA

The Cas9 gene editing system is a very facile but not without limitations and constraints.  For example, the required protospacer-adjacent motif (PAM) constrains target selection when using Cas9-based editing tools, although clever variants with altered PAM requirements complement the original system so circumventing some of these PAM constraints is possible.

Natronobacterium gregoryi is a Halophilic Archaea in the Class Halobacteria and Family Halobacteriaceae.

Gao et al. describe an Argonaute protein from Natronobacterium gregoryi (NgAgo), a haloalkaliphilic archaebacterium that uses a single stranded DNA molecule as a guide.  The guide molecule is 24 bases in length and must be phosphorylated at the 5’ end.

NgAgo can only be loaded with ssDNA guides at 55 C and not at 37 C and once loaded the complex creates double stranded breaks in the target sequence. When NgAgo is being reloaded at 55 C the protein does not have DNA endonuclease activity and activity is only restored when the complex is returned to 37 C.  NgAgo removes several nucleotides in the target region when it cuts DNA although it does not appear to have any exonuclease activity.  Details of the endonuclease cleavage are not yet available.

The high temperature loading requirement might have some interesting implications.  One being that in vivo expression of NgAgo and guideDNAs may not be an option in most organisms as is done in some applications of Cas9.

Argonaute protein plays familiar roles in RNA interference in Eukaryotes but are a diverse family of proteins with interesting properties.

What about efficiency? Although not extensively tested in side by side comparisons, in the data presented NgAgo and a Cas9 directed to the same target had similar efficiencies.

After assessing the ability of some 47 guides targeting 8 genes in the human genome the authors report similar efficiencies among all guides tested – between 21% and 41%.

Mismatches between the guide and target reduced cleavage efficiency and a systematic analysis of  single mismatches at all 24 positions showed that all positions were important with positions 8 -11 being particularly important.  The system seems to be very sensitive to mismatches and this could make off target effects less of a concern.  The authors found that 3 consecutive mismatches eliminated all activity.

The NgAgo system does not depend on anything resembling a protospacer-adjacent motif (PAM), nor does its efficacy decrease when targeting GC-rich regions, as appears to be the case with Cas9.

This is an interesting system and this paper is worth reading and considering.  How this system develops and how much NgAgo might serve as an alternate to Cas9 remains to be seen but it is clear we have not seen the last of technology improvements and diversification in the area of gene editing.


Gao, F., Shen, X. Z., Jiang, F., Wu, Y., Han, C. (2016) DNA-guided genome editing using the Natronobacterium gregoryi Argonaute. Nat Biotech. advance online publication doi: 10.1038/nbt.3547



see : Nature Biotechnology 35, 797 (2017) doi:10.1038/nbt0817-797a



  1. Hi Dave,
    I saw your review on the interesting Argonaute gene modification paper and think I read it differently. To me it looked as though the system would work if you expressed the protein in vivo and supplied the ssDNA in some way as well.
    It wouldn’t work if you tried to load it in vitro though, as some people do with the CRISPR/Cas9 system.
    It seems like a cool system since it would get away from RNA oligos and PAM targets.
    Have I missed something though??

    1. Gareth,

      I would suspect that the protein would work better in vitro, as you could combine the gDNA+Ago at 55˚C, reduce the temperature to 37˚C, end up with a permanently fused ribonucleoprotein, and then transfect that complex.


  2. This review is incorrect about how the loading works. 55 degrees is used to REload the ago, i.e. to remove guides it has already associated with and to reassociate it with new guides. This reloading also partially denatures NgAgo, making it nick rather than fully cleave.
    In vivo loading occurs as the protein is folded, at 37 degrees(this property is the whole reason why the paper is important. We have had DNA guided DNA targeting Agos for years, but they only ever worked at >65˚). If one wanted to use the fully functional protein in vitro as Andre suggested (which I don’t fully agree with the idea of it “working better”, though for HDR you are likely correct as is with Cas9) you would do just what they did in the paper; transfect some 293 cells with your guide and NgAgo plasmid, purify the ago, and there you go, pure protein associated with the guide of your choice. Note that they had to reload only when purifying the protein from bacteria which is because of the presence of endogenous 5’p ssDNA in bacterial cells.

Post a Comment

Your email address will not be published. Required fields are marked *