1. Specificity of RNAi and genome editing
technologies in plants and animals
Brian D. Gregory
Department of Biology
Penn Genome Frontiers Institute
Genomics and Computational Biology Graduate Program
The University of Pennsylvania

2. RNAi in general

3. The active molecules in RNAi are small RNAs (smRNAs)

4. Mechanisms for silencing by smRNAs in plants
and animals

5. Specificity of RNAi is conferred by complementary base-pairing interactions between a smRNA and its target RNA(s)

6. Very few complementary base pairing interactions are required for functional silencing of target RNAs by smRNAs in animals
Witold Filipowicz, Suvendra N. Bhattacharyya & Nahum Sonenberg
Nature Reviews Genetics 9, (February 2008)

7. Very few complementary base pairing interactions are required for functional silencing of target RNAs by smRNAs in animals
Front. Genet., 11 June 2013 | doi: /fgene

8. Plant compared to animal smRNA-mediated silencing and off-target effects
In plants, near perfect complementarity is required for productive silencing of target RNAs. This is likely responsible for the greater target specificity that is observed in plant as compared to animal systems.
However, in large plant genomes (which are common in horticulture and agriculturally important plants) the likelihood of off-target effects increases as a function of genome size.
In plants, most smRNAs silence target RNAs by cleavage not translation inhibition, so usual quite easy to identify off-target effects.
In animals, only seed pairing (and maybe a few 3’ contacts) is required for productive silencing of target RNAs. This results in the greater expected and observed off-target effects observed for this technology when trying to develop RNA-based silencing approaches.
In animals, most smRNAs silence target RNAs through inhibition of translation. Thus, even identifying off-target effects takes significantly more time and effort than for similar experiments focused on determining the answers to this same question in plant systems.
In total, off-target effects are a serious concern with this pathway

9. Too many exogenous smRNAs can also saturate the smRNA biogenesis and function machinery

10. Too many exogenous smRNAs can also saturate the smRNA biogenesis and function machinery
Increase in non-normal smRNAs can lead to exclusion of nomal, endogenous smRNAs from AGO incorporation resulting in developmental defects because of loss of developmentally important smRNA-mediated silencing

11. Nuclease-mediated genome editing

12. Specificity determinants of nuclease-mediated genome editing
Molecular Therapy Nucleic Acids (2012) 1, e3; doi: /mtna

13. CRISPR/CAS-mediated genome editing

14. CRISPR/CAS-mediated genome editing most likely to have off target
effects
Most important for target cleavage

15. Method for detecting off-target effects of nuclease-mediated genome editing

16. Method for detecting off-target effects of nuclease-mediated genome editing

17. Summary of off-target effects as a result of genome engineering by nucleases
Detection of off target effects still in its infancy. Until recently, most commonly used technique was whole genome sequencing to look for new mutations induced by the genome engineering. This is time consuming and labor intensive.
New methods aimed at identifying double-strand breaks (DSBs) (i.e. GUIDE-seq) are coming available and demonstrate that off target effects more prevalent than thought for CRISPR/CAS system. These approaches need to be used now on ZFN and TALEN-based approaches as well, so that comprehensive comparisons can be made.
Early studies suggest that the specific CRISPR used and its targeting region on the genome cause large-scale variation in off target effects. Many more studies needed to truly determine how CRISPRs can be made as specific as possible for all applications.

18. Some broad comparisons of the genomic effects of RNAi/Genome Editing to normal plant breeding techniques
Breeding techniques such as introgression and mutagenesis are likely to result in significantly more genomic changes than targeted RNAi and genome editing approaches
In fact, current breeding techniques (especially introgression and mutagenesis) are currently known to effect more genomic loci and resulting traits than the much more targeted RNAi and genome editing technologies even when considering off-target effects of the latter technologies.
Stable transgene insertion not necessarily required for any of these techniques. However, stable, multi-generation RNAi will require transgenic technologies, and currently most genome editing technologies do involve the use of transgenic approaches.
However, genome editing approaches of desired crop traits is currently being engineered to not rely on stable transgenic insertion.