Example analysis of PASTE

[kclem]

PASTE (courtesy of Yarnall et al.) is a two-step genome editing-assisted approach for integration of large genomic elements. First, an integrase landing site is inserted into the genome via prime editing, and second the integrase delivers the cargo at the desired location.

Here, we use simulated data to show how CRISPResso can be used to measure integrase landing site insertion efficiency.

This is the schematic from figure 1b where the AttB site is inserted using prime editing into ACTB exon 1.

Input parameters

For your reference, I've copied the RNA sequences from the figure and their reverse sequences below:

Extension sequence:
AAGAGCGTCGAGTCCGGCCGAACAGCTGCTGCCGCCAGAGGCAGCAGTCCTAGTAGGCCGGTACCTACTACTATAGCGGCGCGAGCAG
reverse:
GACGAGCGCGGCGATATCATCATCCATGGCCGGATGATCCTGACGACGGAGACCGCCGTCGTCGACAAGCCGGCCTGAGCTGCGAGAA


Nicking sgRNA:
CGTACACGTTCCGGCCGAAG
reverse
GAAGCCGGCCTTGCACATGC

To run CRISPResso using the suite of prime-editing tools, we need to provide:

• -a the WT amplicon sequence: GCGCGCCCGGCTATTCTCGCAGCTCACCATGGATGATGATATCGCCGCGCTCGTCGTCGACAACGGCTCCGGCATGTGCAAGGCCGGCTTCGCGG
• --prime_editing_pegRNA_spacer_seq in RNA 5'->3' order, so for Cas9, the PAM is on the right side: GCTATTCTCGCAGCTCACCA
• --prime_editing_pegRNA_extension_seq RNA 5'->3' order, so in the order of RT>edit>PBS so the reverse of the sequence shown in the figure: GACGAGCGCGGCGATATCATCATCCATGGCCGGATGATCCTGACGACGGAGACCGCCGTCGTCGACAAGCCGGCCTGAGCTGCGAGAA
• prime_editing_pegRNA_scaffold_seq in RNA 5'->3' order. This isn't given in the figure, but a common value is: GTTTTAGAGCTAGAAATAGCAAGTTAAAATAAGGCTAGTCCGTTATCAACTTGAAAAAGTGGCACCGAGTCGGTGC
• --prime_editing_nicking_guide_seq in RNA 5'-3' order, so for Cas9 the PAM is on the right side. This is the reverse of the sequence shown in the figure: GAAGCCGGCCTTGCACATGC

Simulated input

We create the following simulated dataset with 6 reads (let's keep it simple here) and call it input.fq.

@WT1
GCGCGCCCGGCTATTCTCGCAGCTCACCATGGATGATGATATCGCCGCGCTCGTCGTCGACAACGGCTCCGGCATGTGCAAGGCCGGCTTCGCGG
+
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
@WT2
GCGCGCCCGGCTATTCTCGCAGCTCACCATGGATGATGATATCGCCGCGCTCGTCGTCGACAACGGCTCCGGCATGTGCAAGGCCGGCTTCGCGG
+
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
@WT3_plus_mut
GCGCGCCCGGCTATTCTCGCAGCTCACATGGATGATGATATCGCCGCGCTCGTCGTCGACAACGGCTCCGGCATGTGCAAGGCCGGCTTCGCGG
+
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
@PE1
GCGCGCCCGGCTATTCTCGCAGCTCAGGCCGGCTTGTCGACGACGGCGGTCTCCGTCGTCAGGATCATCCGGCCATGGATGATGATATCGCCGCGCTCGTCGTCGACAACGGCTCCGGCATGTGCAAGGCCGGCTTCGCGG
+
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
@PE2
GCGCGCCCGGCTATTCTCGCAGCTCAGGCCGGCTTGTCGACGACGGCGGTCTCCGTCGTCAGGATCATCCGGCCATGGATGATGATATCGCCGCGCTCGTCGTCGACAACGGCTCCGGCATGTGCAAGGCCGGCTTCGCGG
+
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
@SCAFFOLD1
GCGCGCCCGGCTATTCTCGCAGCTCAGGCCGGCTTGTCGACGACGGCGGTCTCCGTCGTCAGGATCATCCGGCCATGGATGATGATATCGCCGCGCTCGTCGCACCGGTCGACAACGGCTCCGGCATGTGCAAGGCCGGCTTCGCGG
+
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA

The first two reads (WT1 and WT2) are wildtype reads with no edits. The third read (WT3_plus_mut) has a deletion at the pegRNA nick site (TCA|CCA > TCA|CA). The fourth and fifth reads (PE1 and PE2) are perfect prime edit insertions of the AttB sequence. And the sixth read (SCAFFOLD1) is the prime-edited insertion plus scaffold sequence (GTC|GTC > GTC|GCACCGGTC).

Running CRISPResso

We run CRISPResso in prime-editing mode using the following command:

CRISPResso -r1 input.fq -a GCGCGCCCGGCTATTCTCGCAGCTCACCATGGATGATGATATCGCCGCGCTCGTCGTCGACAACGGCTCCGGCATGTGCAAGGCCGGCTTCGCGGCGACGATTTTGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGG --prime_editing_pegRNA_spacer_seq GCTATTCTCGCAGCTCACCA --prime_editing_pegRNA_extension_seq GACGAGCGCGGCGATATCATCATCCATGGCCGGATGATCCTGACGACGGAGACCGCCGTCGTCGACAAGCCGGCCTGAGCTGCGAGAA --prime_editing_pegRNA_scaffold_seq GGCACCGAGTCGGTGC --prime_editing_nicking_guide_seq GAAGCCGGCCTTGCACATGC -p 10 --debug

Here, I've added the -p 10 flag to run on 10 processors and the --debug flag to print out a few more things, including the alignment of the pegRNA extension sequence to the WT amplicon sequence:

Alignment between extension sequence and reference sequence:
-------------TTCTCGCAGCTCAGGCCGGCTTGTCGACGACGGCGGTCTCCGTCGTCAGGATCATCCGGCCATGGATGATGATATCGCCGCGCTCGTC-----------------------------------------------
GCGCGCCCGGCTATTCTCGCAGCTCA----------------------------------------------CCATGGATGATGATATCGCCGCGCTCGTCGTCGACAACGGCTCCGGCATGTGCAAGGCCGGCTTCGCGGCGACGAT

Examining CRISPResso output

This produces the correct assignment and quantification of reads:

The experimental layout and analysis outcomes are correctly shown in plot 11a below.

• The location of the PE spacer sgRNA, PE Extension, and PE nicking sgRNA are shown, with three quantification windows (outlined in dotted lines) showing locations where the presence of insertions, deletions, or substitutions, will cause a read to be labeled as 'MODIFIED'. The quantification windows for the nicking sgRNAs are 1bp on each side of the predicted cut site as we expect edits to arise at that location, while the site at the 3' flap excision/repair site is larger (5bp on each site of the excision/repair site, a parameter that can be changed using --prime_editing_pegRNA_extension_quantification_window_size).

• Note that the red bar in the 'PE Extension' annotation at the bottom of the plot signifies that the 'PE Extension' sequence doesn't exactly match the reference sequence in this plot (because it has the AttB landing site there).

• The 1 read with the 1bp deletion at the PE spacer cut site has been identified, and the location of the scaffold incorporation has been identified as well.

The complete report contains more details, but all of the expected results from our simulated dataset can be seen in these two high-level plots. A richer (real) dataset would be more interesting to explore with the full report.

In summary, CRISPResso is a versatile tool, and can be used to analyze complex genome editing designs. Hopefully this example shows how a schematic can be translated to correct parameters for analysis by CRISPResso.

Happy editing!

Kendell