CRISPR in Animal Model Generation - Functionalizing Genome Editing for a Broad Range of Targets

Introduction to CAGE Offerings

Main core offerings from CAGE to the St. Jude's research community include:

• Design, assemble and validate guide RNAs (e.g., RNPs and plasmid)
• Design, assemble and validate guide RNAs (e.g., RNPs and plasmid)
• Create custom clonal cell lines
• Create animal models (e.g., mouse and rat)
• Perform targeted NGS for CRISPR edited samples
• Facilitate high throughput genetic screening by the generation and QC of pooled guide RNA libraries
• Create edited cell pools

Genome editing at CAGE relies on traditional CRISPR editing tools. Briefly, targeting of the Cas9 nuclease to specific genomic sequences depends on the guide RNA and particularly a 20 nucleotide sequence, which interacts with the DNA at the target site. Double strand breaks are induced by Cas9 (e.g., spCas9), which are repaired by the endogenous cell repair pathways (i.e. NHEJ and HDR) resulting in targeted mutagenesis. Depending on the cell repair mechanisms, modifications may include gene disruption (insertions and deletions or indels), direct ligation, deletion, inversion, translocations. Additionally, if the cell is supplied with a donor DNA template, precise single nucleotide modifications or sequence insertions may be achieved.

Using Single Stranded DNA knockin for Animal Model Generation

Single stranded DNA, ssODN (<150 nucleotides) and ssDNA (>150 nucleotides), may be leveraged for knockin insertions. Dr. Pruett-Miller relies on synthetic single stranded DNA, which allows the knockin of larger insertions through direct injection into embryos or electroporation.

As opposed to double stranded DNA or double stranded donor templates, as exemplified by plasmids, ssDNA provides a more efficient strategy for knockin animal model generation. Dr. Pruett-Miller has found that the single strand DNA repair pathway is more efficient, particularly in embryos.

By using this strategy, an integration frequency greater than 50% may be achieved. Additionally, because homology arms in ssDNA donor templates are shorter than those typically used with double stranded plasmid donor constructs, verifying insertions is facilitated by direct targeted NGS analysis (as shown above, with genome and insertion specific primers-Gen/Junc). Dr. Pruett-Miller recommends that insertions should be verified through sequencing from homology arm to homology arm to avoid potential small synthesis errors.

Assaying guide RNA Activity

Several assays may be used in order to determine the activity of guide RNAs to induce double strand breaks. Most methods depend on the delivery of Cas9 and the guide RNA into cells to allow for double strand breaks to take place. Genomic DNA harvested from the cells is PCR amplified, targeting the modification site, which may be analyzed through various downstream assays including T7E1/Cel-1, targeted NGS, TIDE/ICE or IDAA.

Comparison of all the options for downstream assaying guide RNA activity showed that NGS was the best approach to identify active guide RNAs, determine indel sizes and sequences, obtain accurate measure of editing rates for pools of cells and cell clones, and to support high throughput analysis.

Common limitations for T7E1, TIDE, NGS and IDAA assays include PCR bias leading to increased amplification efficiency of small PCR products, inability to detect all chromosomal changes due to loss or sequence changes preventing primer annealing, and lastly amplicon size is limited to <1000 bps due to challenges inherent to genomic DNA amplification. For NGS sequencing approaches, the major limitation is the access to technology, which involves costly instruments and reagents, and requires technical expertise to run the instruments and manage the resulting large data sets.

Genome Editing Workflow with Targeted NGS at CAGE

Briefly, the workflow starts with the introduction of the guide RNAs and Cas9 reagents through transfection or nucleofection for cell modifications and by electroporation or direct injection for animal model generation. Cells are cultured as single cells or as pools. For animal model generation, injected embryos are implanted into a pseudo pregnant mouse. Genomic DNA is harvested and analyzed by targeted NGS to confirm the desired modification with an Illumina platform. For PCR amplification, a two-step PCR method first amplifies the area of interest, and a second PCR reaction adds indexes needed for demultiplexing following sequencing.

Finally, resulting sequencing data containing the information on editing efficiency (e.g., FASTQ), indel size and sequence identity, is analyzed by using CRIS.py.

CRIS.py: What is CRIS.py and How does it Work?

CRIS.py is a python-based program for high-throughput sequence analysis developed by the Pruett-Miller lab, which analyses NGS data for knockout and knockin modifications resulting from gene editing.

Advantage of CRIS.py for NGS analysis:

•   Runs locally in a desktop computer without need of a server
•   Amenable to the analysis of >20 samples
•   Ability to search for >1 donor sequence
•   Generates master data summary that is searchable and sortable

CRIS.py uses the targeted amplified sequence as the reference sequence. Following designation of a start and end sequence (seq_start and seq_end), CRIS.py is able to search through the NGS data set to match sequences having the same seq_start and seq_end. For all matching sequences, CRIS.py measures the length of the sequence and reports the indels in comparison to the reference sequence. For example, a 0bp indel and -1bp indel, would correspond to a sequence with no insertions or deletions and a sequence with one base pair deletion, respectively. For the analysis of knockin modifications, an unlimited number of test sequences may be provided to the program, based on expected editing outcomes, for comparison.

Finally, CRIS.py provides two main outputs: a master summary and a sequence file. The master summary is compiled in a .csv format or excel type file, which is both searchable and sortable, and provides an easy to follow summary of the frequency of indels associated with each gRNA. The sequence file (.txt format), provides the resulting sequences and their frequency. Allowing easy distinction between sequencing errors and true indels. Overall CRIS.py facilitates the gene editing workflow by simplifying the way the data output is presented, which is both concise and intuitive.

The original program and PrettyCRISP.py, an updated version, may be freely used and is accessible at https://github.com/jake-steele/prettyCRIS.py.

Deep sequencing combined with CRIS.py analysis facilitates:

•   Resolving the allele frequency in commonly used cell lines (e.g., N2A).
•   Understanding the heterogeneity of a cell population.
•   Resolving gene dependency in cells.
•   Generating large deletions efficiently.

CRISPR Gene Editing Tips

•   Using bridging oligo (ssODN) increases targeted deletion efficiency.
•   Using single stranded DNA, ssDNA (>200bp) increases knockin efficiency especially in embryos (>50% Knockin efficiency using GenScript long ssDNA service).