Protocol Citation: Jennifer Zevounou, Hicham Belhaj, Chao-Sheng Lo, Estelle Lecluze, Ken Sin Lo, Guillaume Lettre 2025. Perturb-seq Using ABE8e Base Editing to Functionally Characterize a Subset of Blood Pressure- and CAD-Associated Genetic Variants in TeloHAEC Cells. protocols.io https://dx.doi.org/10.17504/protocols.io.dm6gpqpbdlzp/v1
License: This is an open access protocol distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Protocol status: Working
We use this protocol and it's working
Created: June 04, 2025
Last Modified: June 13, 2025
Protocol Integer ID: 219561
Keywords: TeloHAEC, Blood Pressure (BP), Coronary Artery Disease (CAD), Genome-wide association studies (GWAS), Adenine Base Editing (ABE) Screening, scRNA-seq, Pertub-seq, transcriptomic consequences of specific genetic edit, specific genetic edit, pooled adenine base editing, associated genetic variants in telohaec cell, associated genetic variant, genetic variant, adenine base editing, using abe8e base editing, abe8e base editing, transcriptomic consequence, genome, crispr perturbation, human endothelial cell, endothelial cell, coronary artery disease, throughput functional annotation of disease, cell rna, uk biobank data, gwa, blood pressure, sequencing, using uk biobank data
Funders Acknowledgements:
Guillaume Lettre
Grant ID: HG012010
Abstract
This study follows a previously developed pooled Adenine Base Editing (ABE) screen that assessed the functional impact of 1,271 genetic variants associated with blood pressure (BP) traits and coronary artery disease (CAD) in human endothelial cells (TeloHAEC). That screen, based on variants identified through genome-wide association studies (GWAS) and fine-mapped using UK Biobank data, enabled high-throughput functional annotation of disease-linked loci.
Building upon this, the current protocol describes a Perturb-seq approach to characterize a selected subset of these variants at single-cell resolution. This method combines ABE8e-mediated base editing with single-cell RNA sequencing of CRISPR perturbations to elucidate the transcriptomic consequences of specific genetic edits in endothelial cells.
Materials
A
B
C
Name
SKU
VendorName
polybrene (hexadimethrine bromide)
H9268-5G
Merck MilliporeSigma (Sigma-Aldrich)
Tissue culture plate 6 wells Falcon
CA62406-161
VWR International (Avantor)
Falcon tube 50 ml
1443222
Thermo Fisher Scientific
Millipore Steriflip sterile disposable
0.45uM
SE1M003M00
Thermo Fisher Scientific
pMD2.G
12259
addgene
Plus reagent
11514015
Life Technologies
Lipofectamine 2000
11668-019
Life Technologies
BSA Stock Solution
130-091-376
Miltenyi Biotec
Falcon tube 15ml
1495949B
Thermo Fisher Scientific
Opti-MEM
31985070
Life Technologies
psPAX2
12260
addgene
Falcon T-175
83.3912.002
Sarstedt
DMEM High glucose with glutamax and
Sodium pyruvate
All variants were annotated for their presence in:
a coding sequence of gene of interest for endothelial cells (top expressed genes in endothelial cell (Dai et al,PMID: 35610053), differentially expressed gene in teloHAEC +/- TNFa, differentially expressed gene in HAECs depending flow exposure (Wu et al, PMID: 28556776)
an endothelial cell or teloHAEC active +enhancer, characterized with the ENCODE rE2G and scE2G model
Variants were also annotated with their:
finemapping results for Systolic Blood Pressure and Diastolic Blood Pressure (Keaton et al, PMID: 38689001)
base editing results from the previous base editing Screen
Variants selection
Firstly, all variants were filtered based on their BEAN results from the previous Base Editing Screen: variants with an editing rate >= 10% and an effect size, abs(mu_z_adj) > 5, were considered.
Then, variants were filtered based on their relative genes. A few loci were eliminated:
- AIDA : the causal variant that we proposed in Lalonde et al is not tested (it is not an A-target variant)
- ALDH2 : the gene maps to the SH2B3/ATXN2 locus. SH2B3 is very pleiotropic and has a validated role in immune cells.
- CALCRL: the gene is expressed at low levels in teloHAEC. And the putative causal variant (rs880890, PMID: 38602103) is tested and not significant.
- FHL3: expressed at low levels in teloHAEC
- MAT2A: In our PLoS Genet paper, we identified 3 promising variants at this locus. 1 is not screened and the 2 others are not A-to-G variants. Following up on these 5 variants would be difficult to reconcile with our previous results.
At the end, 31 variants present in 6 loci (CDKN1A,ERG,PLPP3,PECAM1,TGFB1,SMAD3) were selected.
Guide RNA (gRNA) design
2
The teloHAEC genome sequence was used to design the gRNA library, with genetic variation in the editing window accounted for when appropriate. Alleles at non-A (or T) variants cannot be installed using the ABE
editing strategy. However, because other A’s may be present in the editing windows and bystander edits could provide useful information, all selected variants were targeted by the gRNA library, independent of the variant genotype in teloHAEC. With bedtools and the teloHAEC genome sequence, the genomic sequence around the targeted variants was retrieved, and five 20-bp gRNAs were designed for each variant, shifting the variant targeted for editing from position #4 in the editing window (gRNA1) to position #5 (gRNA2), #6 (gRNA3), #7 (gRNA4), and #8 (gRNA5). After this initial design, the gRNA sequences were re-aligned to the teloHAEC genome to account for other nearby genetic variants (i.e., variants not associated with BP or CAD but residing near the targeted variants). When appropriate, these variants were introduced into the gRNA sequences to minimize mismatches.
The strategy developed by Ryu et al. (PMID: 38658794) was adopted, adding a reporter sequence for each gRNA. This reporter includes the same 20 nucleotides as the gRNA and is used to measure editing efficiency. In total, 104 gRNAs were designed to target 31 variants.
Pertub-seq library cloning
3
The pooled gRNA library was synthesized by Agilent Technologies.
2- Pool the 10 reactions and purify the PCR product using the QIAquick PCR Purification Kit, following the manufacturer’s instructions. Elute in 20 µL of nuclease-free (ddH₂O) water.
3- Run the PCR product on a 2% agarose gel and excise the 241 bp band. Purify the DNA using the Qiagen Gel Extraction Kit, following the manufacturer’s instructions. Elute in 50 µL of nuclease-free (ddH₂O) water.
4- Perform a second PCR using all the product from PCR1:
PCR2 reaction – Amplify the PCR2 reaction in 10 tubes:
A
B
C
Volume (ul)
10X volume (ul)
PCR1 product from gel extraction
5
50
NEBNext Ultra II Q5 Master Mix (E7805L)
10
100
PCR2 forward primer (10uM)
1
10
PCR2 reverse primer (10uM)
1
10
dd H2O
3
30
Total volume
20
200
Cycling:
A
B
C
D
Temp
Time
Cycles
Step 1
98°
30 sec
1X
Step 2
98°
10 sec
10X
70°
30 sec
72°
30 sec
Step 3
72°
5 min
1X
Step 4
4°
Hold
5- Pool the 10 PCR reactions and purify the product using a PCR purification column, as previously described. Run the purified product on a 2% agarose gel and excise the 299 bp band. Perform gel extraction using the Qiagen Gel Extraction Kit, following the manufacturer's instructions. Elute in 20 µL of nuclease-free (ddH₂O) water.
6
Vector preparation
Restriction Digest of pHKO9-BsmBI library backbone.
1- Add components in the following order:
Volume (ul)
pHKO9-BsmBI (10ug)
15
dd H2O
65
rCutsmart
10
DTT (20mM)
5
BsmBI
5
2- Incubate at 37 °C for 1 hour.
3- Run 1% agarose gel and isolate 8kb band by gel extraction (as described previously).
7
Gibson assembly:
1- Set up the following reaction on ice:
A
B
Volume (ul)
Gibson Assembly Master Mix (2X) (E5510)
10
pHKO9-BsmBI (350ng/ul)
1
PCR2 product from gel extraction
(75ng/ul)
0.7
dd H2O
8.3
Total Volume
20
2- Incubate samples in a thermocycler at 50°C for 1 hour
8
Perform isopropanol precipitation:
Volume (ul)
Gibson reaction mix
20
Isopropanol
20
GlycoBlue
0.4
5M NaCl
0.4
Vortex, incubate at room temperature for 15 min.
Centrifuge at top speed for 15 min.
Carefully remove liquid without disturbing the pellet.
Wash twice with 1 mL of ice-cold 80% EtOH.
Carefully remove all liquid, and air-dry pellet for 2-3 min by keeping the cap open at room temperature.
Add 6 µL ddH₂O and warm at 55°C for 10 min to fully resuspend.
9
Electroporation Protocol
1. Transformation
Add 2 µL of Gibson assembly product (~50 ng) to 25 µL of Lucigen Endura electrocompetent cells.
Electroporate using the following settings:
Cuvette: 1 mm
Capacitance: 10 µF
Resistance: 550 Ω
Voltage: 1700 V
2. Recovery
Immediately add 1 mL of S.O.C. recovery medium to the cuvette.
Transfer to a culture tube and shake at 37 °C for 1 hour.
3. Dilution and Plating
Prepare a 1:1000 dilution by adding 10 µL of the recovered culture to 990 µL of LB.
Plate 100 µL of this dilution (i.e., a 1:10,000 final dilution) onto an agar plate.
Incubate the plate overnight at 37 °C.
4. Library Expansion
Incubate the remaining 990 µL of the undiluted culture in 50 mL of LB broth.
Grow overnight at 37 °C with shaking.
5. Colony Counting & Cloning Efficiency
Count the colonies on the 1:10,000 dilution plate.
Subtract any colonies observed on the negative control plate (no insert) to assess cloning efficiency.
Estimate total colony count by multiplying by the dilution factor.
Goal: Achieve ≥20× coverage of your library.
6. Plasmid Extraction
If coverage requirements are met, proceed with a Midiprep on the overnight bacterial culture.
Sequencing and QC
10
Sequencing was performed at the Centre d’expertise et de services Génome Québec, located at CHU Sainte-Justine. Libraries were sequenced on an Illumina NovaSeq PE150 platform, targeting a read coverage of more than 500 reads per sgRNA per sample.
TeloHAEC Thawing
11
For each vial, prepare a 15 mL tube with 10 mL of warm complete VCBM media supplemented with 0.15 µg/mL puromycin and P/S.
Take a cryovial of cells out of liquid nitrogen storage and thaw in a 37°C water bath until only a tiny bit of ice remains. Vials should contain 1 million cells.
13
Mix cells by gentle shaking, spray off the outside with 70% EtOH, flick down to bring all cells/media to the bottom of the tube, and transfer to a tissue culture (TC) hood.
14
Transfer the content of one vial into a prepared 15 mL tube containing 10 mL of warm complete VCBM.
15
Centrifuge for 5 minutes at 200g, room temperature (RT), and discard the supernatant.
16
Resuspend in 1 mL of complete VCBM, then count and assess viability.
17
Plate cells in a T75 flask with 15 mL of complete media.
Thaw TeloHAEC cells as described above and allow them to recover for a few days in complete VCBM. Throughout the recovery period, ensure that cells are maintained below 80% confluence.
34
Before infection, remove the virus stock from the -80 °C freezer and thaw it slowly on ice in a biosafety level 2 (BL2) culture room.
35
Trypsinize and count TeloHAEC cells. Based on the cell count, distribute 2 × 10⁵ cells per well into six wells of a 6-well plate. Ensure that the cell suspension volume per well is ≤ 800 µL.
Additionally, prepare one extra control well that will not undergo selection.
Tissue culture plate 6 wells FalconVWR International (Avantor)Catalog #CA62406-161
36
Example of volumes of cells, medium, virus and polybrene needed for a virus titration in a 6-well plate
The next day, passage the cells into a larger culture dish and replace the medium with VCBM complete media supplemented with penicillin/streptomycin (p/s) and 8 µg/mL puromycin to begin selection. For the control well (no virus), maintain cells in regular VCBM complete media without puromycin.
It is crucial to retain all cells or split them equally across all conditions, as this step is essential for accurate calculation of the multiplicity of infection (M.O.I.).
The puromycin concentration (8 µg/mL) was established by performing a kill curve on TeloHAEC cells.
38
Selection duration is defined as the number of days required to completely eliminate all cells in the control well (without virus); in this case, it is 6 days.
After the selection period:
Trypsinize the cells and perform a cell count.
Plot a curve with:
X-axis: Volume of virus used
Y-axis: Number of cells that survived selection
3. Use the resulting curve to determine the volume (or approximate volume) of virus that results in ~30% cell survival, relative to the unselected control well. This volume of virus, when used to infect 2 × 10⁵ cells, corresponds to an MOI of 0.3. Scale up accordingly if a larger number of cells is to be infected.
Generate TeloHAEC-SFFV-ABE8e cell line
39
The TeloHAEC model expressing dCas9-SFFV-ABE8e-Blast was generated via lentiviral transduction (as described above) using the pLenti-SFFV-ABE8e-(D10A) SpRY-P2A-Blast vector. Transduction of wild-type TeloHAEC cells (TeloHAEC-WT) was performed to establish a stable cell line expressing the ABE8e base editor. Transduced cells were subjected to selection with 5 µg/mL blasticidin for 7 days. Expression of the ABE8e-(D10A) SpRY-P2A-Blast protein in the resulting population was subsequently confirmed by Western blot analysis.
The maintenance medium for the TeloHAEC-SFFV-ABE8e cell line consists of Complete VCBM supplemented with penicillin/streptomycin, 0.15 µg/mL puromycin, and 2.5 µg/mL blasticidin.
Pooled ABE screen for scRNA-seq
40
TeloHAEC-SFFV-ABE8e cells were thawed from liquid nitrogen storage and passaged twice prior to infection, as described above. Briefly, cells were seeded into four T225 culture flasks, each containing 5 million cells in 26.75 mL of VCBM supplemented with 25 µL of polybrene (1 mg/mL). Infection was performed using the volume of virus corresponding to an MOI of 0.3, as determined by prior titration. These four flasks were prepared to allow for subsequent TNF-α treatment (10 ng/mL for 4 hours) in two flasks, with the remaining two flasks serving as untreated controls.
41
After 24 hours of infection, cells were subjected to selection with 8 µg/mL puromycin for 14 days in complete VCBM supplemented with penicillin/streptomycin (p/s). Throughout the selection period, cultures were maintained below 80% confluence.
42
At the end of the selection period, two flasks of cells were treated with 10 ng/mL of TNF-α for 4 hours, while two additional flasks were kept as untreated controls. Following treatment, cells were trypsinized, counted, centrifuged, and resuspended in PBS containing 0.04% BSA to achieve a final concentration of 1,500 cells/µL.
43
Cells were then loaded into the 10x Chromium instrument following the manufacturer’s protocol (Chromium GEM-X Single Cell 5' Reagent Kits v3 with Feature Barcode technology for CRISPR screening). A total of four lanes on the 10x chip were loaded, corresponding to the four conditions (two controls and two TNF-α–treated samples). Each lane was loaded with 29,000 cells, targeting a recovery of 20,000 cells per condition.
44
After GEM generation, reverse transcription, library construction, and quality control (QC) were performed according to the manufacturer’s instructions.
Sequencing
45
Eight libraries were generated using the 10x Genomics protocol described above: four gene expression libraries, one for each condition, and four corresponding CRISPR libraries capturing the gRNAs. The libraries were pooled such that the gene expression libraries received four times more sequencing reads than the CRISPR libraries.
Sequencing was carried out at the Centre d’expertise et de services Génome Québec, located at CHU Sainte-Justine, using an Illumina NovaSeq PE100 platform. A total of 3,200 million reads were generated, targeting a read depth of over 20,000 reads per cell for the gene expression libraries and 5,000 reads per cell for the CRISPR libraries.