Multiplexed target enrichment and high-resolution NGS analysis of on- and off-target CRISPR-Cas9 editing events via rhAmpSeq technology Michael A Collingwood*, Mollie Schubert, Garrett R Rettig, Gavin Kurgan, Matthew McNeill, Junzhou Wang, Stephanie Fiedler, Christopher Vakulskas, and Mark A Behlke Integrated DNA Technologies, Coralville, IA, USA
* Corresponding author: mcollingwood@idtdna.com
Introduction
rhAmpSeq quantification of editing B. AAVS1-T2
10
10
80
On-target editing: WT Cas9 = 32.3% Alt-R HiFi Cas9 = 84.3%
70 60 50 40 30 20
1
63 50 37 24 11 -2 -15 -28 -41 -54 -67 -80 -93 -106 -119 -132
0 1
Amplicon position (0 = Cas9 cut site) 0.1
Cas9 targeted sites 40
A.
Off-target score
rhAmp primer
10 uM gRNA
Activation by DNA polymerase
100
% INDEL formation
87 rhAmp PCR amplicon
Individual 1.5X0.75 SPRIXpurification Individual SPRI purification
Table 2. GUIDEseq identification of off-target sites in HEK-293–Cas9 cells with Alt-R crRNA:tracrRNA complexes
AAVS1-670 RNP delivery HEK-293 cells
gRNA delivery HEK-293–Cas9 cells
Activation of rhAmp primers by Rnase H2 cleavage
53
CCTCTAAGGTTTGCTTACGA
Editing site
rhAmp primer
71
TAAGCAAACCTTAGAGGTTC
C.
Targeted rhAmp PCR
22
GGACGCACCATTCTCACAAA
AAVS1-670
On-target editing: 10 µM gRNA = 17.0% 10 µM gRNA + tag = 10.0%
Cas9 targeted sites
GGGGCCACTAGGGACAGGAT
AAVS1-692
4 uM HiFi RNP
100
0.1
20-nt guide sequence (5′→3′)
AAVS1-292
90
10
Table 1. Guide RNAs targeting the AAVS1 locus
AAVS1-T2
4 uM WT RNP
10 uM gRNA + GuideSeq
100
AAVS1 gRNA survey Guide name
RNP delivery HEK-293 cells
gRNA delivery HEK-293–Cas9 cells 10 uM gRNA
100
INDEL frequency
Table 3. Unbiased detection of off-target sites informs rhAmpSeq design criteria
Editing event characterization AAVS1-T2 D. On-target edit width
Index PCR amplicons with P7 & P5 universal primers
% On-target editing
10
10 uM gRNA + GuideSeq
On-target editing: 10 µM gRNA = 99.2% 10 µM gRNA + tag = 99.2%
4 uM WT RNP 100
10
1
1
0.1
0.1
Cas9 targeted sites
% Off-target editing
4 uM HiFi RNP
INDEL frequency
•
rhAmpSeq methodology
CRISPR-Cas9 facilitates genome editing using a targeting guide RNA (gRNA) that directs the Cas9-RNA complex to a genomic DNA locus; however, this prokaryotic nuclease system is not always well adapted to maintain high specificity when targeting much larger eukaryotic genomes. There is a significant need for a streamlined pathway to identify, quantify, and characterize the total picture of both on- and off-target genome editing events. Here, a solution is described to first use established techniques to identify double-stranded breaks (DSBs) for a given genome editing experiment and second accurately quantify editing efficiency at each edited locus using a multiplexed, targeted enrichment approach called rhAmpSeq technology.
% INDEL formation
•
rhAmpSeq panel application
AAVS1-T2 On-target INDEL profile
35 30 25 20 15 10 5 0 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10
On-target editing: WT Cas9 = 99.2% Alt-R HiFi Cas9 = 99.2%
INDEL size Total editing
Frameshift edits
Cas9 targeted sites % Insertions % Deletions
Guide
AAVS1-T2
AAVS1-292
AAVS1-692
AAVS1-670
On-target editing (% of total reads)
7.4
97.9
94.3
99.8
Number of off-target sites detected
297
8
8
4
Tables 1–2. Positive prediction of improved Cas9 gRNA specificity in the AAVS1 locus and confirmation of specificity using GUIDEseq. Literature precedence for gRNA AAVS1-T2 led to its common use as a gRNA to target AAVS1, a “safe-harbor” locus for transgene integration in human cells. However, the off-target homology and predictive score suggest better guides could be used. (Table 1) Alternative gRNAs targeting the AAVS1 locus were selected to have improved predicted specificity using the IDT gRNA checker (www.idtdna.com/Cas9checker). (Table 2) The 4 gRNAs were compared in GUIDEseq [1] experiments to validate on- and off-target cleavage sites in an unbiased fashion via Nucleofection of 10 µM Alt-R crRNA:tracrRNA complexes (IDT) into HEK-293 cells with stable expression of Cas9. Data analysis was as published [2] with improved local alignment capability.
Table 3, Figure 1. Complete rhAmpSeq workflow from primer design through data analysis. (Table 3) Using the rhAmpSeq Custom Design Tool (IDT), rhAmp primers were computationally designed to flank each identified Cas9 cleavage site. rhAmp primers were synthesized, analyzed by ESI-MS, pooled, and used in target enrichment via multiplex amplification in a single reaction. (Figure 1A) In the rhAmp PCR methodology, RNaseH2 recognizes and cleaves the rhAmp primers via substrate recognition of the RNA base within the DNA:RNA duplex and removes a 3′-blocking group from perfectly hybridized primers [3]. Activated primers are extended to generate locusspecific amplicons, enriching validated and/or predicted Cas9 target sites, thus eliminating primer dimers or mis-amplification. A high-throughput protocol has been optimized for time-savings at critical points in the protocol (indicated by ). Compared to the standard protocol, the high-throughput method eliminates costly SPRI clean-up of individual samples and enables completion of library preparation in <1 day. (Figures 1B–C) rhAmpSeq panels for AAVS1-T2 (111-plex) and AAVS1-670 (9-plex) enable detection of all targets of interest in a single reaction, providing valuable time- and cost-savings, while enabling experiments testing multiple variables (i.e., Cas9 source, gRNA modifications, cell lines, RNP dose, HiFi vs. WT Cas9, etc.). Each experiment is sequenced by Illumina sequencing at an exceedingly high read map rate and target uniformity (see Fig 2). (C) rhAmpSeq technology confirms specificity of editing at only the AAVS1 locus is >99% for Alt-R HiFi Cas9 RNP delivery of AAVS1-670. (B) Delivery of the low-specificity gRNA is likewise dramatically improved to >84% on-target accuracy using Alt-R HiFi Cas9 RNP electroporation into HEK-293 cells. (Figure 1D) Panel D shows the complete characterization of editing events available for all targets for the AAVS1-T2 gRNA.
Summary
• rhAmpSeq panel uniformity B.
282-plex percent mapped 10 ng gDNA
Percent mapped
100%
Mean: 99.7%
15 ng gDNA Mean: 99.9%
25 ng gDNA
50 ng gDNA
Mean: 99.9%
Mean: 99.9%
95%
90%
85% NA12878
NA24385
NA12878
NA24385
NA12878
Coriell gDNA samples
NA24385
NA12878
NA24385
282-plex target coverage uniformity 10 ng gDNA
≥0.2X Mean target coverage
A.
% In-frame edits % Frameshifted edits
15 ng gDNA
25 ng gDNA
•
50 ng gDNA
100%
•
95% 90% 85% 80%
Novel rhAmpSeq technology facilitated an accurate evaluation and determination of the most specific Cas9 gRNA option that targets the AAVS1 locus. The percent mapped and target uniformity displayed (Figure 2) for the custom 282-plex rhAmpSeq panel demonstrates reliability of the rhAmpSeq technology, which is built upon the primer design tool as well as the RNaseH2-containing rhAmpSeq master mix. Absent in silico gRNA screening to alleviate off-target-editing, rhAmpSeq technology analysis of genome-wide editing by Alt-R HiFi Cas9 indicates that this enzyme dramatically reduces off-target editing and maintains excellent on-target performance [4].
• The multiplex enrichment and data analysis provide an efficient workflow
75%
from primer design through sequencing, enabling accurate and comprehensive characterization of editing events.
70% 65% NA12878
NA24385
NA12878
NA24385
NA12878
NA24385
NA12878
NA24385
References
Coriell gDNA samples
Figure 2. Custom 282-plex rhAmpSeq pool gives very high mapping rate and uniformity. The uniformity of rhAmpSeq technology amplification was tested using a 282-plex of primer pairs and 2 sources of Coriell gDNA at 10, 15, 25, and 50 ng input gDNA amounts. (A) “Percent-mapped” is defined as the number of merged reads mapping to the human genome over the total number of reads. For this custom panel of 282 primer pairs, the mean percent-mapped is >99% for all conditions tested (n = 48 for each gDNA source and amount). (B) The same custom panel and libraries generated from the various Coriell samples were assessed for uniformity of target amplification. Uniformity is measured by the fraction of targets that have a read depth that is >0.2X of the mean read depth. The requirement of the custom panel is that >85% of all targets meet this metric. For both percent mapped and target uniformity measurements, the resultant read counts are from Illumina sequencing (MiSeq® system, 2 x 150 paired-end reads, v2 chemistry).
1. Tsai SQ, Zheng Z, et al. (2015) Nat Biotechnol. 33(2):187–197. 2. Tsai SQ, Topkar VV, et al. (2016) Nat Biotechnol. 34(5):483. 3. Dobosy JR, Rose SD, et al. (2011) BMC Biotechnol. 11:80. 4. Vakulskas CA, Dever DP, et al. (2018) Nat Medicine. 24(8):1216–1224. For research use only. © 2019 Integrated DNA Technologies, Inc. All rights reserved. Alt-R and rhAmp are trademarks of Integrated DNA Technologies, Inc., and are registered in the USA. rhAmpSeq is a trademark of Integrated DNA Technologies, Inc. All other marks are the property of their respective owners. For specific trademark and licensing information, see www.idtdna.com/trademarks.