Nov 21, 2025
Single Cell Transcriptional Perturbome in Pluripotent Stem Cell Models: Protocols 1-4
- elisa.balmas 1
- 1University of Torino
Protocol Citation: elisa.balmas 2025. Single Cell Transcriptional Perturbome in Pluripotent Stem Cell Models: Protocols 1-4. protocols.io https://dx.doi.org/10.17504/protocols.io.e6nvw4q29lmk/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: October 18, 2025
Last Modified: November 21, 2025
Protocol Integer ID: 230162
Keywords: 10X genomics, iPS2-10X-seq, iPS2-CITE-seq, iPS2-MULTI-seq, seq from 10x genomic, conjunction with the appropriate 10x genomics protocol, appropriate 10x genomics protocol, use of cell multiplexing oligo, cell multiplexing oligo, parallel to the standard gene expression library, cell transcriptome, pluripotent stem cell model, 10x genomic, seq relies on microfluidics partitioning, gene expression kit, standard gene expression library, 10x genomics single cell multiome atac, single cell transcriptional perturbome, gene expression library, unique cell, cell doublet, shrna barcode library library, absence of additional multiplexing barcode, cell number, microfluidics chip, single microfluidics channel, microfluidics partitioning, additional multiplexing barcode, shared unique cell, multiplexed sample, manufacturer for barcoded sample pool, treated cell, seq rely, silico matching of uci, barcoded sample pool, based scrna
Funders Acknowledgements:
Giovanni Armenise-Harvard Foundation
Grant ID: Career Development Award 2021
ERC starting grant
Grant ID: 101076026
Disclaimer
Contact [email protected] and or [email protected] for further tips and assistance
Abstract
iPS2-10X-seq relies on microfluidics partitioning-based scRNA-seq from 10X Genomics. Despite the higher costs, the ease and reproducibility of this commercial method have led to its widespread adoption. The procedure integrates the manufacturer’s protocol to generate an additional NGS library containing UCI-BCs. This is sequenced in parallel to the standard gene expression library to enable in silico matching of UCI-BCs and single-cell transcriptomes through their shared unique cell barcodes. Of note, iPS2-10X-seq provides built-in support for the cost-saving strategy of reasonably overloading the microfluidics chip, since cell doublets are readily identified by the co-expression of two different UCI-BCs and filtered out as part of the standard catcheR_10Xcatch pipeline. Thus, it is possible to load the cell number recommended by the manufacturer for barcoded sample pools even in the absence of additional multiplexing barcodes. Other cost-saving strategies include the use of cell multiplexing oligos (CMOs) or barcoded antibodies against housekeeping cell surface antigens to multiplex tet-untreated and tet-treated cells in a single microfluidics channel, enabling robust control versus knockdown experimental design without having to perform two separate reactions.
This Protocol describes the following procedures:
- Prepare the gene expression library
- Prepare the shRNA barcode library
- Library pooling and NGS
This protocol was optimized using the Chromium Next GEM Single Cell 3’ Reagent Kits v3.1 (Dual Index) from 10X Genomics and has been successfully validated with the next-generation GEM-X Single Cell 3’ Reagent Kits v4, without requiring changes to primer sequences. It must be used in conjunction with the appropriate 10X Genomics protocols: CG00315 Rev E (single sample), CG000388 Rev B (multiplexed samples), or CG000419 Rev D (high-throughput) for v3.1;, and CG000731 Rev A for v4. Notably, v4 no longer supports cell multiplexing oligo (CMO)-based on sample multiplexing but remains fully compatible with antibody-derived tag (ADT)–based
strategies.
A first variation, iPS2-CITE-seq, enables surface protein profiling and ADT-based multiplexing using barcoded TotalSeq-B antibodies from BioLegend. When using antibodies, refer to protocol CG000149 Rev D (antibody labeling) in combination with CG000317 Rev E for v3.1 or CG000731 Rev A for v4.
A second variation, iPS2-multi-seq, adapts the protocol for use with the 10X Genomics Single Cell Multiome ATAC & Gene Expression kit. It should be paired with CG000365 Rev C (nuclei isolation) and CG000338 Rev F (library construction).
Image Attribution
Figure SP3.1
Materials
- NEBNext High-Fidelity 2X PCR Master Mix
- iPS2-10X-seq_true5 (10 µM)
- iPS2-10X-seq_inner (10 µM)
- Preamplified 10X Genomics cDNA (50 ng)
- Nuclease-free water
- Dual Index Kit TT Set A primers
- Qubit dsDNA High Sensitivity kit
Troubleshooting
Protocol 1: Generation of iPS2-seq plasmids and hiPSCs
The iPS2-seq pipeline starts with a two-step cloning procedure for a barcoded pool of shRNAs into an AAVS1 locus targeting plasmid to generate all-in-one tet inducible cassette.
This begins with designing single stranded DNA (ssDNA) oligonucleotides each carrying an shRNA with a polythymidine (pT) Pol III terminator sequence, a uniquely matched barcode (BC), and a random sequence serving as a unique clonal identifier (UCI).
The shRNA and UCI-BC sequences are separated by a short multicloning site (SalI, SwaI, AscI) that is needed at a later stage, and are flanked by short regions of homology to the destination vector to enable assembly.
The oligos are synthesized, pooled, and converted to double stranded DNA (dsDNA) through a single round of isothermal primer extension from the 3’ end using a mesophilic DNA polymerase with strong strand displacement activity.
This is essential to overcome the secondary structure of shRNAs in the oligos and minimize representation biases due to differential hairpin strength.
dsDNA fragments are then cloned through Gibson assembly into an AAVS1 targeting plasmid downstream of an inducible H1 Pol III promoter and upstream of a pA signal.
The vector contains a puromycin (puro) resistance gene trap and is based on our previously reported plasmid, but was modified to remove repetitive sequences that led to recombination errors in the next cloning step.
The plasmid also includes a novel thymidine kinase (TK) cassette outside of the AAVS1 homology arms. This allows subsequent negative selection of cells carrying random, off-target integrations of the whole plasmid as they become sensitive to fialuridine (FIAU), a thymidine analog preferable to ganciclovir.
The resulting intermediate plasmids are then modified in pool to insert a CAG-OPtettR cDNA cassette between the shRNA and UCI-BC sequence.
This second cloning step reconstitutes the tet-ON system while marking the 3’ UTR of the OPtettR with the UCI-BC sequence matched to a given shRNA.
This is performed via restriction digestion and ligation, as the CAG promoter is too GC-rich for efficient and reliable PCR amplification.
An optional strategy built into the system allows depletion of the intermediate plasmid from the final population through restriction digestion with SwaI.
The final plasmids are utilized to engineer pools of hPSCs enriched for clones expressing single inducible shRNAs.
Plasmids are transfected alongside vectors encoding for obligate heterodimer zinc finger nucleases (ZFNs) to induce double-strand breaks specifically at the AAVS1 locus and facilitate homology-directed repair (HDR) of the inducible barcoded shRNA cassette.
A second AAVS1 targeting plasmid is also co-delivered: this is a filler vector that carries no relevant cargo but has a neomycin (neo) resistance gene trap that allows enrichment of cells that integrated an shRNA on one AAVS1 allele and the filler sequence on the other allele; these biallelic edited cells are co-selected with puro and neo.
Additional negative selection with FIAU kills clones that carry randomly integrated shRNAs.
Collectively, this genome editing strategy is designed to reduce the frequency of double shRNA integrations and maximise the fraction of cells expressing a single shRNA from the AAVS1 locus.
Of note, the first cloning step can result in a low rate of swap between the shRNA and its expected BC based on oligonucleotide design and synthesis.
This issue likely arises from DNA polymerase template switching during dsDNA synthesis and/or Gibson assembly.
The problem can be efficiently countered by next generation sequencing (NGS) analysis of Illumina-compatible libraries spanning the shRNA and UCI-BC region from the intermediate plasmid pool; this enables re-assignment of swapped shRNAs to the observed barcode, leveraging the fact that each plasmid also carries a UCI (also essential for tracking hPSC clones downstream).
Plan the screening
Before initiating a screening experiment, it is essential to define three key elements: screening objectives, available budget, and overall feasibility.
The flow chart below summarizes key parameters for optimizing molecular cloning, genome editing, and sequencing efficiency, followed by a worked example: a screen targeting 18 genes with 5 shRNAs per gene plus 10% control constructs.
Figure SP1.1. Screening design example
Flow chart illustrating how to estimate the number of cells required for sufficient iPS2-seq coverage, accounting for dropout rates from cloning to transfection and single-cell sequencing. The top section defines the screening goal and provides an example estimation based on our protocol; the lower section shows a worked example with 100 shRNAs. scRNA-seq calculations are based the maximum loading capacity per reaction for either the Chromium Next GEM Single Cell 3’ HT GEM Kit v3.1 or the Chromium GEM-X Universal 3’ v4 Kit.
Effective screening design should begin with:
Target gene selection — Define the genes to perturb and decide how many shRNAs per gene will be used. We recommend a minimum of 3 shRNAs per gene, and ideally 5, to buffer against off-target effects or poorly performing constructs. Include at least 10% non-targeting control shRNAs.
Cell editing scale — Estimate the number of cells to genome-edit in order to achieve robust clonal representation for downstream iPS2-10X-seq analysis. An example calculation is provided in Figure SP1.1, accounting for both bacterial clone yield and the number of hiPSC clones required for sequencing.
Sequencing depth — Determine how many cells need to be profiled to capture biological variability and achieve statistical power. Depending on experimental goals and budget, controls may include non-targeting shRNAs (e.g., B2M, SCR), tet-untreated cells, or both. We strongly recommend including tet-untreated cells to control for clonal variability.
Design shRNA oligonucleotides
ssDNA oligonucleotides for iPS2-seq can be designed using catcheR_design as described in Supplemental Protocol 4 - Oligonucleotides design. This function automatically performs the following steps, reducing the chance of manual errors when designing multiple oligos:
- Identify validated shRNAs or design appropriate sequences. In our lab we use the Broad Institute TRC library, available at the GPP Web Portal. When using this resource take the ~58 bp forward oligos and remove the sequences required for cloning in pLKO vectors (typically the first 4 nucleotides and the last 6 nucleotides).
- If an shRNA does not start with A or a G, add an initial G (required for efficient transcription by the H1 Pol III promoter).
- Construct the cloning oligos by adding the following sequences to each shRNA, in order:
- At the 5’, the Gibson homology region 5’-AGTTCCCTATCAGTGATAGAGATCCC-3’
- At the 3’, the polythymidine (pT) Pol III terminator sequence 5’-TTTTTTTT-3’
- At the 3’, SalI, SwaI, and AscI restriction sites 5’-GTCGACATTTAAATGGCGCGCC-3’
- At the 3’, a random sequence (UCI; i.e., 5’-NNNNNN-3’)
- At the 3’, an shRNA-specific barcode (BC; i.e., 5’-CAGTTCCA-3’)
- At the 3’, the Gibson homology region 5’-GTAGCTGCGTGATCAGC-3’
The resulting oligos should look like Figure SP1.2
Note: These sequences are passed as arguments to catcheR_design. The recommended minimum length for UCIs and BCs is 6 and 8 bp, respectively. Tools such as BARCOSEL can be used to generate and select BCs with optimal nucleotide distribution.
Figure SP1.2. Example of iPS2-seq ssDNA oligo
Annotated SMAD2.1 shRNA sequence extracted from TRC ID:TRCN0000010477 and modified by
catcheR_design.
- Modify or replace any cloning oligo that contains BglII (5’-AGATCT-3’) or MluI (5’-ACGCGT-3’) restriction sites that would interfere with subsequent cloning steps
- Synthesize oligos as high quality ssDNA. In our lab we use the oPools Oligo Pools service from Integrated DNA Technologies
Pooled cloning step 1: obtain the intermediate plasmid pool
This section outlines the pooled conversion of barcoded shRNA-containing ssDNA oligos into dsDNA by second-strand synthesis (SSS), and the insertion of such fragments in pAAV-Puro_siKD2.0.
- Prepare the oligo pool:
(a) Spin down the lyophilized ssDNA and resuspend it at 50 μM in nuclease-free water
(b) Dilute the 50 μM stock 1:10 in nuclease-free water to make a 5 μM working stock
2. Convert ssDNA to dsDNA using Bst3.0 DNA polymerase
(a) Denature and snap cool the ssDNA oligo pool
i. Mix the following for 8 reactions in a single tube on ice:
| A | B | C | D | |
| Reagent | one reaction (µL) | eight reactions (µL) | Final conc. | |
| ssDNA oligo pool (5µM) | 5 | 40 | 0.5nM | |
| siKDbc_REV Primer (10µM) | 2.5 | 20 | 0.5μM | |
| Deoxynucleotide (dNTP) Solution Mix (10mM) | 2 | 16 | 400μM | |
| Nuclease-free water | 15.5 | 124 | ||
| Total volume | 25 | 200 |
Table SP1.1. SSS reaction mix 1
ii. Distribute 25 μL of the reaction mix into 8 PCR tubes
iii. Place PCR tubes in a thermocycler and run the following cycling conditions:
| A | B | C | D | |
| Steps | Temperature | Time | Cycles | |
| Denaturation | 98°C | 5 min | 1 | |
| Hold (rapid ramp down) | 4°C | Hold |
Table SP1.2. Cycling conditions for ssDNA denaturation and snap cooling
(b) Perform SSS
i. Mix the following for 8 reactions in a single tube on ice:
| A | B | C | D | |
| Reagent | One reaction (µL) | Eight reactions (µL) | Final conc. | |
| MgSO4 (100mM) | 1 | 8 | 2mM | |
| Isothermal Amplification Buffer II (10X) | 5 | 40 | 1X | |
| Bst 3.0 DNA Polymerase (8 U/µL) | 2 | 16 | 0.32 U/µL | |
| Nuclease-free water | 17 | 136 | ||
| Reaction from previous step | 25 | 200 | ||
| Total volume | 25 | 200 |
Table SP1.3. SSS reaction mix 2
ii. Add 25 μL of this reaction mix to each of the 8 PCR tubes from step 2(a)iii
iii. Place PCR tubes in a thermocycler and run the following cycling conditions:
| A | B | C | D | |
| Steps | Temperature | Time | Cycles | |
| Annealing | 55°C | 5 min | 1 | |
| Extension | 72°C | 15 min | 1 | |
| Inactivation | 80°C | 5 min | 1 | |
| Slow ramp down | 75°C | -0.1deg/s | 1 | |
| Hold | 75°C | 4 min | 1 | |
| Slow ramp down | 70°C | -0.1deg/s | 1 | |
| Hold | 70°C | 4 min | 1 | |
| Slow ramp down | 10°C | -0.1deg/s | 1 | |
| Hold | 10°C | Hold |
Table SP1.4. Cycling conditions for SSS
(c) Remove residual ssDNA
i. Add 4 μL of exonuclease I (20 U/μL) to each PCR tube and pipette mix
ii. Place PCR tubes in a thermocycler and run the following cycling conditions:
| A | B | C | D | |
| Steps | Temperature | Time | Cycles | |
| Digestion | 37°C | 30 min | 1 | |
| Inactivation | 80°C | 20 min | 1 | |
| Slow ramp down | 75°C | -0.1deg/s | 1 | |
| Hold | 75°C | 4 min | 1 | |
| Slow ramp down | 70°C | -0.1deg/s | 1 | |
| Hold | 70°C | 4 min | 1 | |
| Slow ramp down | 10°C | -0.1deg/s | 1 | |
| Hold | 10°C | Hold |
Table SP1.5. Cycling conditions for ssDNA removal
(d) Clean up and concentrate the dsDNA
i. Pool all 8 reactions into a single tube
ii. Column purify using QIAquick PCR Purification Kit, following its manual
iii. Elute in 30 μL elution buffer
Note: the dsDNA shRNA pool can be stored at -20 °C for at least one week
Prepare the plasmid backbone
(a) Prepare a digestion mix and an undigested control mix following the table below:
| A | B | C | D | |
| Reagent | Digestion reaction (µL) | Undigested control (µL) | Final conc. | |
| pAAV-Puro_siKD2.0 plasmid | Variable | Variable | 5µg | |
| FastDigest Green Buffer (10X) | 15 | 1.5 | 1X | |
| FastDigest MluI | 5 | 0 | ||
| FastDigest BglII | 5 | 0 | ||
| FastAP (1 U/μL) | 5 | 0.5 | 0.03 U/μL | |
| Nuclease-free water | Top up to 150 | Top up to 15 | ||
| Total volume | 150 | 15 |
Table SP1.6. pAAV-Puro_siKD2.0 restriction digestion mix
(b) Incubate at 37 °C overnight (~16 h), followed by 10 min at 80 °C for heat inactivation
Perform electrophoretic separation of the insert (step 2(d)iii) and backbone (step 3b) using 2% and 0.8% (w/v) agarose gels in TBE, respectively. Cast gels including SYBR Safe DNA gel stain to visualize DNA, preferably using a blue light transilluminator (avoid UV light if possible). Figure SP1.3 shows exemplary electrophoreses
(a) Use 50 bp and 1 kbp DNA ladders for the insert and backbone, respectively
(b) Run 10% of each sample in a separate well and use this to monitor the electrophoretic separation, so as to avoid unnecessary exposure to high-energy light of the gel extracted DNA (particularly important if using UV light by necessity)
(c) Use FastDigest Green Buffer (10X) as loading dye for the insert: commonly used DNA loading dyes containing bromophenol blue and/or xylene cyanol FF can interfere with visualization of the small shRNA pool fragment
(d) If a smear is present in the insert, continue electrophoresis until the 135 bp band is clearly separated
(e) Run 90% of the digested empty plasmid (~140 µL) in a single well as this will facilitate gel extraction: merge multiple wells using autoclave tape if needed
(f) Load the plasmid undigested control next to 10% of the digestion reaction and continue the electrophoretic separation until the 8715 bp backbone band is completely separated from the undigested control (this can take several hours: use a low voltage and/or change buffer to prevent gel overheating). Incomplete separation can result in contamination of undigested or partially digested plasmid, which can dramatically increase transformation background
Figure SP1.3. Electrophoretic separation and gel extraction during cloning step 1 (A) Separation of shRNA pool insert (SSS); a PCR-amplified fragment indicates the expected molecular weight. (B) Separation of digested plasmid backbone.
Use a clean scalpel to excise the DNA bands corresponding to the shRNA pool insert (135 bp) and plasmid backbone (~8.7 kbp) and place them in a microcentrifuge tube
Proceed with gel extraction using the QIAEX II Gel Extraction Kit following the manual eluting in 20 µL EB; quantify the DNA concentration using NanoDrop
Setup the desired assembly alongside control reactions following the table below:
| A | B | C | D | E | |
| Reagent | Reaction (µL) | BB only CTRL (µL) | No Assembly CTRL (µL) | Final conc. | |
| Gel extracted shRNA fragment (135bp) | Variable | 0 | 0 | 14.34ng (172.4 fmol) | |
| Gel extracted iPS2-seq backbone (8715bp) | Variable | Variable | Variable | 50ng (9.32 fmol) | |
| NEBuilder HiFi DNA Assembly Master Mix (2X) | 10 | 10 | 0 | 1X | |
| Nuclease-free water | To 20 | To 20 | To 20 | ||
| Total volume | 20 | 20 | 20 |
Table SP1.7. Pooled cloning step 1 assembly mix
Incubate the reactions at 50 °C for 15 min, then transfer to ice
Note: avoid longer incubation times as they can increase background
Transform 2 µL of assembly or control reactions in 50 µL NEB 5-alpha competent E. coli (high-efficiency), following the manufacturer’s instructions
Plate transformed bacteria
Pre-warm LB-agar plates containing 100 µg/mL ampicillin to 37 °C
To facilitate estimation of colony number and cloning efficiency, plate 10% of transformed bacteria on one dish, and the remaining 90% on a second dish
Incubate the plates at 37 °C overnight
Estimate the number of bacterial clones by counting the colonies in the plate containing 10% of the transformation of the desired assembly (>200 are to be expected)
Notes:
The number of colonies in the matching backbone-only control and no assembly control plates should be at least 10-fold and 100-fold lower; it is advisable to repeat the procedure should a substantially higher background be observed
Bacterial plates can be stored at 4 °C for up to one week to allow for the subsequent optional quality controls
OPTIONAL. Confirm cloning efficiency by bacterial colony PCR
(a) For each colony to be screened, aliquot in PCR plates a 15 µL PCR mix based on nuclease-free water and containing: iPS2-seq_step1QC_F Primer (200 nM), iPS2seq_seq_QC_R Primer (200 nM), dNTPs (200 µM), MgCl2 (2 mM), Colorless GoTaq Flexi buffer (1X), and GoTaq DNA polymerase (0.025 U/µL)
(b) Pick an individual bacterial colony from the 10% plate, mix it by pipetting in 10 µL nuclease-free water aliquoted in a PCR plate, and transfer 5 µL of the suspension to the PCR mix for a final volume of 20 µL. The remaining 5 µL can be transferred to a PCR plate containing LB with 100 µg/mL ampicillin and stored at 4 °C for later use. Repeat for all colonies
(c) Place the PCR plate in a thermocycler and run following cycling conditions: (1) 95 °C for 5 min; (2) 95 °C for 30 sec; (3) 60 °C for 30 sec; (4) 72 °C for 1 min; (5) go to step 2 for 34 times; (6) 72 °C for 2 min; (7) hold at 10 °C
(d) Aliquot 5 µL of each colony PCR in a new plate, add 1 µL of 6X DNA loading dye, and perform an electrophoretic run on a 1.5% (w/v) agarose gel in TBE
Note: clones with the correct insert deliver a PCR product of 543 bp.
OPTIONAL. Confirm cloning accuracy by Sanger sequencing
Aliquot 5 µL of each colony PCR in a new plate and add 1.5 µL of a mastermix containing 0.5 µL of exonuclease 1 (20 U/µL) and 1 µL of shrimp alkaline phosphatase
(1 U/µL)
Incubate the reaction at 37 °C for 30 min, followed by heat inactivation for 15 min at 80 °C
Add 5 μL of iPS2-seq_QC_R primer (10 μM) to each well, and run Sanger sequencing.
Note: alignment of each electropherogram to the common sequence 5’-TTTTTTTGTCGACATTTAAATGGCGCGCCNNNNNNNNNNNNNNGTAGCTCGCTGATCAGC-3’
simplifies theextraction of the BC (last 8 Ns); the electropherogram can then be aligned
to the expected oligo based on the BC, to verify the correct sequence of the associated shRNA
Collect all bacterial colonies from the 90% plate by washing and scraping the surface of LB-agar plates with 5 mL of LB broth with 100 µg/mL ampicillin
Inoculate the bacterial suspension in a clean Erlenmeyer flask with 45 mL LB broth containing 100 µg/mL ampicillin, and grow the bacterial cultures at 37 °C at 225 rpm for 16 h
Note: inoculate bacteria in a sterile environment and use a sterile flask to minimize the chances of contamination. Do not grow bacteria for more than 16 h
Isolate the intermediate plasmid pool from bacterial cultures using the QIAGEN plasmid midiprep kit following the manual, and quantify the DNA concentration using NanoDrop
Pooled cloning step 2: obtain the final plasmid pool
This section outlines the generation of the final plasmid pool by re-inserting the CAG-OPTtetR into the intermediate plasmid pool.
Prepare the following restriction digestion reaction of the empty pAAV-Puro_siKD2.0 plasmid with SalI and MluI to extract the CAG-OPTtetR insert:
| A | B | C | D | |
| Reagent | Digestion reaction (µL) | Undigested control (µL) | Final conc. | |
| pAAV-Puro_siKD2.0 plasmid | Variable | Variable | 5µg | |
| FastDigest Green Buffer (10X) | 15 | 1.5 | 1X | |
| FastDigest MluI | 5 | 0 | ||
| FastDigest SalI | 5 | 0 | ||
| Nuclease-free water | Top up to 100 | Top up to 10 | ||
| Total volume | 100 | 10 |
Table SP1.8. pAAV-Puro_siKD2.0 second restriction digestion mix
Prepare the following restriction digestion reaction of the intermediate plasmid pool from step 15 of the previous section with SalI and AscI to obtain the backbone:
| A | B | C | D | |
| Reagent | Digestion reaction (µL) | Undigested control (µL) | Final conc. | |
| Intermediate iPS2-seq plasmid | Variable | Variable | 5µg | |
| FastDigest Green Buffer (10X) | 15 | 1.5 | 1X | |
| FastDigest SalI | 5 | 0 | ||
| FastDigest SgsI (AscI) | 5 | 0 | ||
| FastAP (1 U/μL) | 5 | 0.5 | 0.03 U/μL | |
| Nuclease-free water | To 150 | To 15 | ||
| Total volume | 150 | 15 |
Table SP1.9. Intermediate plasmid pool restriction digestion mix
Incubate at 37 °C overnight (~16 h), followed by 10 min at 80 °C for heat inactivation.
Perform electrophoretic separation of all fragments using 0.8% (w/v) agarose gel in TBE supplemented with SYBR Safe. Figure SP1.4 shows an exemplary electrophoresis.
Note: follow the same general recommendations described for step 7 of the previous section, paying particular attention to separating the backbone from the undigested plasmid.
Figure SP1.4. Electrophoretic separation and gel extraction during cloning step 2. Separation of the CAG-OPTtetR insert and two digested intermediate plasmid pool backbones.
Use a clean scalpel to excise the ~8.8 kbp backbone from the digested intermediate plasmid pool and the ~2.4 kbp CAG-OPTtetR insert from pAAV-Puro_siKD2.0.
Proceed with gel extraction using the QIAEX II Gel Extraction Kit following the manual and eluting in 20 µL EB; quantify the DNA concentrations using NanoDrop.
Set up desired ligation alongside a control reaction following the table below:
| A | B | C | D | |
| Reagent | Reaction (µL) | BB only CTRL (µL) | Final conc. | |
| Gel extracted CAG-OPTtetR fragment (2392bp) | Variable | 0 | 81.59ng (55.38 fmol) | |
| Gel extracted intermediate iPS2-seq pool (8795bp) | Variable | 1 | 100ng (18.46 fmol) | |
| Rapid Ligation Buffer (5X) | 4 | 4 | 1X | |
| T4 DNA Ligase (5 U/μL) | 1 | 1 | 0.25 U/μL | |
| Nuclease-free water | To 20 | To 20 | ||
| Total volume | 20 | 20 |
Table SP1.10. Pooled cloning step 2 ligation mix
Incubate the reaction at RT for 1 h.
Transform 5 µL of ligation and control reaction in 50 µL NEB 5-alpha competent E. coli (high-efficiency), following the manufacturer’s instructions.
Plate and grow overnight transformed bacteria according to the general recommendations described for step 9 of the previous section, particularly regarding splitting the bacteria on two plates (10% and 90%).
Estimate the number of bacterial clones by counting the colonies in the plate containing 10% of the transformation of the desired assembly ( 3e20 are to be expected).
Notes:
(a) the number of colonies in the backbone-only control should be at least 5-fold lower.
(b) bacterial plates can be stored at 4 °C for up to one week.
OPTIONAL. Confirm cloning efficiency by bacterial colony PCR
Perform colony PCRs and electrophoresis as described for step 14 of the previous section, except using primers iPS2-seqstep2QCF and iPS2-seqQCR
Notes:
i. Clones with the correct insert deliver a PCR product of 312 bp; intermediate plasmids are not amplified, while the parental pAAV-Puro_siKD2.0 plasmid gives a 295 bp band (Figure S1C)
ii. Sanger sequencing can also be performed as described for step 15 of the previous section, to confirm the distribution of shRNA BCs; however, at this final stage, the shRNA and barcode cannot be co-amplified
Collect all bacterial colonies and inoculate them as described for steps 16-17 of the previous section. Be extra careful to avoid contamination at this stage.
Isolate the final plasmid pool from bacterial cultures using the QIAGEN plasmid midiprep kit with endotoxin-free buffers, and quantify the DNA concentration using Qubit.
Perform a diagnostic restriction digestion using EcoRI, MluI, and SwaI: the molecular weight of the expected bands is 5669 bp and 5518 bp
Note: additional bands at 2002 bp or 4343 bp would indicate contamination with intermediate plasmids or pAAV-Puro_siKD2.0, respectively. These contaminants can be removed by digestion with SwaI or BglII and MluI, respectively, followed by re-transformation
OPTIONAL: prepare NGS libraries for quality control of plasmid pools
This section outlines the optional generation of Illumina-compatible next-generation sequencing (NGS) libraries to determine the composition of intermediate and final plasmid pools (Figure SP1.5A, Figure SP1.6A). Data analysis relies on catcheRstep1QC and catcheRstep2QC, respectively, as described in Supplemental Protocol 4 - Pooled cloning step 1 QC and Pooled cloning step 2 and hiPSC genome editing QC.
QC n1: Intermediate plasmid pool QC
PCR #1: amplify the shRNA and UCI-BC region from the intermediate plasmid pool while adding the TruSeq read 2 adapter, a diversity index (DI, a random 12 bp sequence that ensures adequate clustering during NGS and can be leveraged to partially demultiplex reads to eliminate duplicates from the subsequent PCR), and partial TruSeq read 1 adapter
(a) Prepare a PCR reaction according to the following calculations:
| A | B | C | |
| Reagent | Volume (µL) | Final conc. | |
| Intermediate iPS2-seq plasmid pool | Variable | 150 ng | |
| ThermoPol Reaction Buffer (10X) | 2.5 | 1X | |
| Deoxynucleotide (dNTP) Solution Mix (10mM) | 0.5 | 200μM | |
| siKDbc_FW Primer (10µM) | 1.5 | 0.5μM | |
| siKDbc_REV Primer (10µM) | 1.5 | 0.5μM | |
| MgSO4 (100mM) | 1 | 4mM | |
| Deep Vent DNA Polymerase (2 U/µL) | 0.25 | 0.02 U/µL | |
| Nuclease-free water | To 25 | ||
| Total volume | 25 |
Table SP1.11. Intermediate plasmid pool QC, PCR #1 mix
(b) Place the PCR tube in a thermocycler and run the following cycling conditions:
| A | B | C | D | |
| Steps | Temperature | Time | Cycles | |
| Initial Denaturation | 95°C | 2 min | 1 | |
| Denaturation | 95°C | 30 sec | 3 | |
| Annealing | 54-59°C (increase by 2.5°C each cycle) | 15 sec | ||
| Extension | 72°C | 30 sec | ||
| Denaturation | 95°C | 30 sec | 9 | |
| Annealing | 63°C | 15 sec | ||
| Extension | 72°C | 30 sec | ||
| Final extension | 72°C | 5 min | 1 | |
| Hold (rapid ramp down) | 4°C | Hold |
Table SP1.12. Cycling conditions for PCR #1, intermediate plasmid pool QC
Note: a "touch-up" strategy minimizes the annealing of shRNA hairpins during the second part of the reaction: once a sufficient number of molecules have integrated the TruSeq adapters, they anneal to the primers at higher temperatures
(c) Add 25 µL of nuclease-free water to the reaction and perform a double-sided size selection using 0.6X & 1.2X volumes of SPRIselect beads, following the manufacturer’s instructions and eluting in 20 µL of elution buffer
(d) Run 2 µL of the undiluted PCR product on a TapeStation High Sensitivity D1000 Screen Tape assay, following the manual. Expect a single peak of ~203 bp. An exemplary result is reported in (Figure SP1.5B)
(e) Quantify the DNA concentration in the range 100-400 bp, and dilute it to 1 ng/µL
PCR #2: prepare dual-indexed Illumina TruSeq sequencing libraries
(a) Prepare a PCR reaction according to the following calculations:
| A | B | C | |
| Reagent | Volume (µL) | Final conc. | |
| PCR#1 product (1 ng/ul) | 2 | 2 ng | |
| ThermoPol Reaction Buffer (10X) | 2 | 1X | |
| Deoxynucleotide (dNTP) Solution Mix (10mM) | 0.4 | 200μM | |
| Dual Index primers Kit TT Set A | 4,00 | ||
| MgSO4 (100mM) | 0.8 | 4mM | |
| Deep Vent DNA Polymerase (2 U/µL) | 0.2 | 0.02 U/µL | |
| Nuclease-free water | 10.6 | ||
| Total volume | 20 |
Table SP1.13. Intermediate plasmid pool QC, PCR #2 mix
(b) Place the PCR tube in a thermocycler and run the following cycling conditions:
| A | B | C | D | |
| Steps | Temperature | Time | Cycles | |
| Initial Denaturation | 95°C | 2 min | 1 | |
| Denaturation | 95°C | 30 sec | 3 | |
| Annealing | 45-47°C (increase by 2.5°C each cycle) | 15 sec | ||
| Extension | 72°C | 45 sec | ||
| Denaturation | 95°C | 30 sec | 9 | |
| Annealing | 63°C | 15 sec | ||
| Extension | 72°C | 30 sec | ||
| Final extension | 72°C | 5 min | 1 | |
| Hold (rapid ramp down) | 4°C | Hold |
Table SP1.14. Cycling conditions for PCR #2, intermediate plasmid pool QC
Add 30 µL of nuclease-free water to the reaction and perform a double-sided size selection using 0.65X & 0.85X volumes of SPRIselect beads, following the manufacturer’s instructions and eluting in 20 µL of elution buffer
Run 2 µL of the undiluted PCR product on a TapeStation High Sensitivity D1000 Screen Tape assay, following the manual. Expect a single peak of ~287 bp
Figure SP1.5. NGS library generation from intermediate plasmid pools
(A) Library structure and primer sequences used for PCR #1 and PCR #2; read 1 is also indicated.
(B) Exemplary TapeStation High Sensitivity D1000 Screen Tape assays of size-selected DNA from PCRs #1 and #2
QC n2: Final plasmid pool QC
PCR #1: amplify UCI-BCs from the final plasmid pool, adding TruSeq adapters and DIs
| A | B | C | |
| Reagent | Volume (µL) | Final conc. | |
| Final iPS2-seq plasmid pool | Variable | 150 ng | |
| Q5 Reaction Buffer (5X) | 5 | 1X | |
| Deoxynucleotide (dNTP) Solution Mix (10mM) | 0.5 | 200μM | |
| iPS2-seq_step2NGS_Primer (10 µM) | 1.25 | 0.5μM | |
| iPS2-seq_NGS_R Primer (10 µM) | 1.25 | 0.5μM | |
| Q5 Hot Start High-Fidelity DNA Polymerase (2 U/µL) | 0.25 | 0.02 U/µL | |
| Nuclease-free water | To 25 | ||
| Total volume | 25 |
Table SP1.15. Final plasmid pool QC, PCR #1 mix
| A | B | C | D | |
| Steps | Temperature | Time | Cycles | |
| Initial Denaturation | 98°C | 30 sec | 1 | |
| Denaturation | 98°C | 10 sec | 9 cycles | |
| Annealing | 67°C | 30 sec | ||
| Extension | 72°C | 20 sec | ||
| Final extension | 72°C | 2 min | 1 | |
| Hold (rapid ramp down) | 4°C | Hold |
Table SP1.16. Cycling conditions for PCR #1
Purify, quantify, and dilute the PCR product following steps 35.2, 35.3, and 35.4 of the previous section, except for using 0.6X & 1.5X volumes of SPRIselect beads for double-sided size selection. Expect a single peak of ~144 bp (Figure SP1.6B)
PCR #2: prepare dual indexed Illumina TruSeq sequencing libraries
| A | B | C | |
| Reagent | Volume (µL) | Final conc. | |
| PCR#1 product (1 ng/ul) | 1 | 1 ng | |
| Dual Index primers Kit TT Set A | 4,00 | ||
| NEBNext high fidelity 2X PCR master mix | 10 | 1X | |
| Nuclease-free water | 5,00 | ||
| Total volume | 20 |
Table SP1.17. Final plasmid pool QC, PCR #2 mix
| A | B | C | D | |
| Steps | Temperature | Time | Cycles | |
| Initial Denaturation | 98°C | 30 sec | 1 | |
| Denaturation | 98°C | 10 sec | 8 cycles | |
| Annealing | 58°C | 20 sec | ||
| Extension | 72°C | 20 sec | ||
| Final extension | 72°C | 30 sec | 1 | |
| Hold (rapid ramp down) | 4°C | Hold |
Table SP1.18. Cycling conditions for PCR #2
Purify and quantify the PCR product following steps 36.3 and 36.4 of the previous section, except for using 0.65X & 1.5X volumes of SPRIselect beads for double-sided size selection. Expect a single peak of ~228 bp (Figure SP1.6B)
Perform NGS using NextSeq 1000/2000 and 50 cycles reagents, with the following settings: read 1 - 68 cycles; index 1 - 10 cycles; index 2 - 10 cycles (>10,000 reads/shRNA)
Figure SP1.6. NGS library generation from final plasmid pools
(A) Library structure and primer sequences used for PCR #1 and PCR #2; read 1 is also indicated.
(B) Exemplary TapeStation High Sensitivity D1000 Screen Tape assays of size-selected DNA from PCRs #1 and #2.
Genome editing: obtain hPSC clonal pools
This section outlines the generation of hPSCs genome edited at the AAVS1 locus with the
final iPS2-seq plasmid pool, and the optional genotyping of individual clones as quality control
(Figure SP1.1).
Preparatory steps:
Determine the number of nucleofections required to obtain the total target number of hPSC clones, based on the number of shRNAs, the target number of clones per shRNA, and the efficiency of genome editing
Notes:
i. The protocol describes plasmid delivery by nucleofection as this is most efficient, but if the necessary instrumentation is not available, plasmids can instead be transfected as described in the STAR METHODS
ii. While our standard protocol uses ZFNs to target the AAVS1 locus, alternative programmable nucleases such as TALENs and CRISPR/Cas9 can also be used.
We opted for ZFNs because we previously demonstrated >95% on-target integration efficiency in hPSCs. That said, our targeting vector is fully compatible with other systems, including TALENs and widely used CRISPR/Cas9 reagents (e.g., Addgene #59025, #59026, #129726). Besides plasmid delivery, in side-by-side comparisons, we observed comparable integration efficiencies using 200,000 cells per nucleocuvette strip co-electroporated with the iPS2-seq vector and either 1 µg ZFN plasmids or Cas9 RNPs (1 µg Cas9 protein + 200 ng sgRNA)
iii. Perform a pilot genome editing experiment to empirically determine the efficiency of genome editing in an hPSC line used for iPS2-seq for the first time
Prepare the following plasmids using the QIAGEN plasmid midiprep kit with endotoxin-free buffers, diluting them to 1 µg/µL (2.5 μg/nucleofection):
i. Pool of pAAV-Puro_siKD2.0 with the barcoded shRNAs (step 32 from Pooled cloning step 2: obtain the final plasmid pool)
ii. pAAV-Neo_CAG
iii. pZFN-AAVS1_ELD
iv. pZFN-AAVS1_KKR
Prepare Geltrex-coated 24-well culture dishes (1 well per nucleofection)
Culture hPSCs in Essential 8 at ~40% confluency (1 x 10^6/nucleofection)
16 h before nucleofection, refresh the hPSC culture media to Essential 8 supplemented with 2 µM Thiazovivin. Do not add antibiotics until step 11
Prepare hPSC Essential 8 supplemented with CEPT [50 nM Chroman 1; 5 µM Emricasan; 0.1% Polyamine Supplement; 0.7 µM trans-ISRIB] (1.5 mL/nucleofection)
Remove the coating solution from the pre-coated culture dish, add 1 mL of Essential 8 with CEPT per well, and place the dish in the incubator to acclimatize to 37 °C
Prepare a mix of equal mass of the four plasmids from step 1b (10 µg/transfection) and nucleofection buffers (18 µL P3 Supplement 26 82 µL Nucleofector Solution per reaction)
Obtain a single cell suspension by treating hPSCs with StemPro Accutase, following the manufacturer’s instructions, aliquot 1 x 10^6^ hPSCs/nucleofection in a conical tube, and pellet the cells at 100 g for 5 min
Remove the supernatant, gently resuspend the pellet with the nucleofection mix from step 5, transfer 110 µL/nucleofection in a Nucleocuvette Vessel (avoiding bubbles), place the cuvette into the 4D-Nucleofector System and run program CA137
Add 500 µL of Essential 8 with CEPT to each cuvette and use the suction pipette to gently transfer the cell suspension to the 24-well plate from step 4. Place in the incubator
24 h post nucleofection, refresh media with Essential 8, and prepare Geltrex-coated 100 mm dishes (1 per nucleofection)
48 h post nucleofection, dissociate cells as small clumps with 0.5 mM EDTA in DPBS, and replate in Essential 8 with 2 µM Thiazovivin (using one 100 mm dish per well of 24-wp)
Refresh media daily with Essential 8. Antibiotics can be added from now on
When hPSCs are ~50% confluent (~5-6 days post-nucleofection), perform dual positive drug selection for 4 days by supplementing Essential 8 with 0.5 µg/mL puromycin and 25 µg/mL Geneticin, refreshing media daily; include 2 µM Thiazovivin for the first 2 days
Note: to optimize the procedure for a specific hPSC line, perform a kill curve with 0.25-2 µg/mL puromycin and 12.5-100 µg/mL Geneticin to identify the minimal concentration of each drug that in combination, eliminates all unedited hPSCs within 72 h
When hPSCs colonies are ready for passaging (~1-2 mm in diameter), count the number of colonies (clones), dissociate cells as small clumps with 0.5 mM EDTA in DPBS, and replate in Essential 8 with 2 µM Thiazovivin at a density of ~1 clones per cm^2^
From 24 h post passaging until 5 days post passaging, refresh media daily with Essential 8 supplemented with 200 nM fialuridine (FIAU) to perform the negative selection
hPSC clonal pools are now ready for screening experiments and/or cryopreservation
Notes:
(a) Minimize the number of passages before the induction of gene knockdown, to minimize the expansion of hPSC clones with a genetic/epigenetic growth advantage
(b) To account for the possible emergence of hPSC clones with neuroectodermal bias (“iPSC-neuro” clones), ensure sufficient clonal representation per perturbation. iPS2-seq enables robust retrospective control for this source of variability; however, it may still be helpful to screen for expression of early neuroectodermal markers such as ZIC1 or UNC5D. If a given pool contains an overrepresentation of iPSC-neuro clones, consider repeating the targeting or depleting biased clones using, for example, a UNC5D antibody for FACS or MACS
(c) For hPSC growth and differentiation protocols involving media containing FBS and/or BSA, batch-test these reagents for the absence of tetracycline contamination (i.e., by generating a clonal line expressing a validated inducible shRNA, to confirm that shRNA expression is not leaky in the absence of exogenously added tetracycline)
(d) To induce knockdown, add 1 ng/mL tetracycline, refreshing the media daily or every 48 h, when compatible for the specific differentiation protocol
hiPSC pool QC
This procedure aims to determine clonal diversity and shRNA representation in the genome-edited hiPSC pool by generating Illumina-compatible NGS libraries of UCI-BCs that can be analyzed using catcheR_step2QC as described in Supplemental Protocol 4 - Pooled cloning step 2 and hiPSC genome editing QC.
After hPSC passaging, collect the cells from a confluent 6-well plate, and pellet the cells at 100 g for 5 min.
Note: Cell pellets can be frozen at -20 °C if needed.
Obtain purified genomic DNA (gDNA) of the genome-edited hiPSCs with the Monarch Spin gDNA Extraction Kit.
Elute samples with 50-100 µL of elution buffer and assess DNA purity and quantity by NanoDrop and Qubit broad range dsDNA.
Use 500 ng of gDNA per PCR reaction, performing as many reactions as needed to achieve coverage of at least 100-fold the estimated number of clones.
Note: One nanogram of human gDNA contains approximately 152 genome equivalents; therefore, 500 ng corresponds to ~76,000 potential UCI-BC templates at the AAVS1 locus.
Using the same primers as the Final plasmid pool QC, perform a PCR to amplify UCI-BCs from the genomic DNA, adding TruSeq adapters and DI, according to the adjusted calculations and cycling conditions:
| A | B | C | |
| Reagent | Volume (µL) | Final conc. | |
| gDNA iPS2-seq pool | Variable | 500 ng | |
| Q5 Reaction Buffer (5X) | 10 | 1X | |
| Deoxynucleotide (dNTP) Solution Mix (10mM) | 1 | 200μM | |
| iPS2-seq_step2NGS_Primer (10 µM) | 2.5 | 0.5μM | |
| iPS2-seq_NGS_R Primer (10 µM) | 2.5 | 0.5μM | |
| Q5 Hot Start High-Fidelity DNA Polymerase (2 U/µL) | 1,00 | 0.04 U/µL | |
| Nuclease-free water | To 50 | ||
| Total volume | 50 |
Table SP1.19. PS2-seq transfected cells QC, PCR #1 mix
| A | B | C | D | |
| Steps | Temperature | Time | Cycles | |
| Initial Denaturation | 98°C | 30 sec | 1 | |
| Denaturation | 98°C | 10 sec | 26 cycles | |
| Annealing | 64°C | 30 sec | ||
| Extension | 72°C | 20 sec | ||
| Final extension | 72°C | 2 min | 1 | |
| Hold (rapid ramp down) | 10°C | Hold |
Table SP1.20. Cycling conditions for PCR #1
Purify the expected ~144 bp amplification product and purify it using SPRIselect beads with a double side selection using 0.6X & 1.5X bead ratios to remove primer dimers and small fragments.
Dilute the PCR product to 0.1 ng/µL in EB buffer and perform indexing PCR according to Table SP1.17 and Table SP1.18.
An indexed library product of ~228 bp is expected. Purify the reaction using a double-sided SPRIselect bead cleanup (0.65X & 1.5X ratios).
Run 2 µL of the undiluted and 1:10 final library on a TapeStation High Sensitivity D1000 Screen Tape assay, following the manual.
Note: This is the same amplicon described for the final plasmid QC (Figure SP1.6B), but adjusted to work on genomic DNA material.
Perform NGS using MiSeq or NextSeq 1000/2000 with the following settings: read 1 - 68 cycles; index 1 - 10 cycles; index 2 - 10 cycles (>10,000 reads/clone).
OPTIONAL: hiPSC clone genotyping
This optional procedure can be used to assess the efficiency of biallelic AAVS1 genome editing in a subset of hPSC clones, or to isolate homozygously edited clones for validation studies. Genotyping leverages various genomic PCRs to determine on-site integration, copy number, and random integrations (Figure SP1.7)
Figure SP1.7. iPS2-seq clone genotyping strategy
Schematic of the genotyping strategies to monitor integration of pAAV-Puro_siKD2.0 in allele 1 and pAAV-Neo_CAG in allele 2. Primer binding sites and directions are shown on the plasmid or integrated cassettes. PCR types and products are numbered as in Table SP1.23.
After step 13 of Genome editing: obtain hPSC clonal pools, pick and expand individual hPSC clones.
Note: treat clones with 200 nM FIAU to test for random integrations.
Prepare a culture dish with 2-5 x 105 cells, and extract gDNA using the Monarch Spin gDNA Extraction Kit, diluting to a concentration of 50-200 ng/µL
Perform PCR reactions according to the following calculations and cycling conditions, employing the primer pairs listed in Table SP1.23. Include the following controls: wild type hPSCs; pAAV-Puro_siKD2.0; no template
| A | B | C | |
| Reagent | one reaction (µL) | Final conc. | |
| Genomic DNA | 1 | 50-200ng | |
| LongAmp Taq Reaction Buffer (5X) | 2 | 1X | |
| Deoxynucleotide (dNTP) Solution Mix (10mM) | 0,3 | 300μM | |
| Forward Primer (10µM) | 0,5 | 0.5μM | |
| Forward Primer (10µM) | 0,5 | 0.5μM | |
| DMSO | 0,2 | 2\% | |
| LongAmp Taq DNA Polymerase (2.5U/µL) | 0,4 | 0.1 U/µL | |
| Nuclease-free water | To 10 | ||
| Total volume | 10 |
Table SP1.21. Genotyping mix
| A | B | C | D | |
| Steps | Temperature | Time | Cycles | |
| Initial Denaturation | 94°C | 5 min | 1 | |
| Denaturation | 94°C | 15 sec | 35 cycles | |
| Annealing | Tab step SP1.23 | 30 sec | ||
| Extension | 65°C | Tab step SP1.23 | ||
| Final extension | 65°C | 5 min | 1 | |
| Hold (rapid ramp down) | 4°C | Hold |
Table SP1.22. Cycling conditions for genotyping
Run half the PCR products on 0.8% (w/v) agarose gels and determine the genotype based on Table SP1.24
Table SP1.23. Genotyping strategies for iPS2-seq clone genotyping
aResult of PCR on wild type AAVS1 allele
bResult of PCR on iPS2-seq-targeted AAVS1 allele
cResult of PCR on pAAV-Puro_siKD2.0 (positive control for off-target plasmid integration)
dVariable parameter in PCR protocol (Table SP1.22)
Table SP1.24. Interpretation of iPS2-seq clone genotyping
Protocol 2: iPS2-sci-seq library preparation
Decoding single-cell transcriptional perturbomes in iPS2-seq relies on detecting the shRNA-associated UCI-BC within the 3’ UTR of the OPTtetR transgene (Figure 1A). This can be in principle achieved with any scRNA-seq method that tags and counts the 3’ ends of mRNAs, usually as a result of pA-primed RT.
iPS2-sci-seq is a home-brew adaptation of the 2-level indexing sci-RNA-seq protocol19, that relies on 2 rounds of transcriptome barcoding through sequential pool-split of defined cell numbers so that the resulting combinatorial barcode in each cell is statistically most likely unique. This approach requires limited custom reagents and supports the analysis of 1,000-10,000 cells per experiment, a scale compatible with the everyday needs of most laboratories. Minor adaptations of this protocol that require additional reagents can enable the 3-level indexing variant of sci-RNA-seq (sci-RNA-seq3), which supports analysis of over ~1 million cells per experiment. Since transcript counts in sci-RNA-seq are notoriously low and zero-inflated (i.e., genes may not be counted at least once in cells that express them), it is key to incorporate a strategy that explicitly enriches for sequences containing the UCI-BC, thereby detecting it with higher sensitivity and specificity. Therefore, in the iPS2-sci-seq protocol, UCI-BCs are enriched at both transcriptome barcoding steps: first during RT, using a set of target-specific primers that share RT barcodes with standard pT primers, well by well, and secondly during PCR, using a set of target-specific primers that share i7 barcodes with standard P7 primers, once again well by well. The resulting multiplexed NGS library pools contain both single-cell transcriptomes and UCI-BCs, which share library structure and carry identical combinatorial indexes, cell by cell.
This Supplemental Protocol describes the following procedures:
1. Reagent setup
2. Nuclei preparation
3. Reverse transcription
4. Nuclei sorting
5. Second-strand synthesis
6. Tagmentation
7. Indexing PCR and NGS
Reagent setup: Prepare the RT and PCR oligos using nuclease-free water:
Obtain master stocks of the 96 indexed iPS2-sci-seq_pT primers (250 µM) and iPS2-sci-seq_tetR primers (100 µM). Use the RT_indexes indicated in Table SP2.1
Prepare a working stock of multiplexed RT primers by diluting in the same plate iPS2-sci-seq_pT primers (25 µM) and iPS2-sci-seq_tetR primers (2.5 µM), matching well positions so that pooled primers share the same RT index
Obtain master stocks of indexed iPS2-sci-seq_P5, iPS2-sci-seq_P7, and iPS2-sci-seq_P7A (all at 100 μM). Use the i5 and i7 indexes indicated in Table SP2.1.
Note: it is not necessary to obtain 96 indexed oligos of each type, as i5 and i7 can be combined to generate unique combinations.
Prepare a working stock of iPS2-sci-seq_P5 (10 μM).
Prepare a working stock of multiplexed P7 primers by diluting in the same plate iPS2-sci-seqP7 (10 μM) and iPS2-sci-seqP7A (5 μM), matching well positions so that pooled primers share the same i7 index.
Prepare the following buffers using nuclease-free water, and chill them on ice:
Nuclei buffer (store at 4 °C for up to 1 month, 5 mL/sample): 10 mM Tris-HCl, pH 7.4, 10 mM NaCl, 3 mM MgCl2.
Nuclei lysis buffer (made fresh, 1 mL/sample): nuclei buffer with 0.1-0.3% IGEPAL CA-630, 1% SUPERase In RNase Inhibitor (20 U/μL) and 1% BSA (20 mg/mL).
Note: the optimal concentration of IGEPAL CA-630 for a given cell type should be determined in a pilot experiment. For reference, 0.1% works for hPSCs, but not for hPSC-CMs which are lysed appropriately only at 0.3%.
Nuclei suspension buffer (make fresh, 3 mL/sample): nuclei buffer with 1% SUPERase In RNase Inhibitor (20 U/μL) and 1% BSA (20 mg/mL).
4% PFA (store at 4 °C for up to 1 month, 10 mL/sample).4 °C
DPBS without Ca2+ and Mg2+ (store at 4 °C for up to 1 month, 10 mL/sample).4 °C
| Well | RT_index (5'-3') | i7 for primer (5'-3') | i7 for sample sheet (5'-3') | i5 for primer & sample sheet, forward strand (5'-3') | i5 for sample sheet, reverse strand (5'-3') | Well | RT_index (5'-3') | i7 for primer (5'-3') | i7 for sample sheet (5'-3') | i5 for primer & sample sheet, forward strand (5'-3') | i5 for sample sheet, reverse strand (5'-3') | ||
| A01 | TTCTCGCATG | CCGAATCCGA | TCGGATTCGG | CTCCATCGAG | CTCGATGGAG | E01 | AGAGAAGGTT | TTCGTTCCAT | ATGGAACGAA | AGGCGAGAGC | GCTCTCGCCT | ||
| A02 | TCCTACCAGT | ATAAGCCGGA | TCCGGCTTAT | TTGGTAGTCG | CGACTACCAA | E02 | CATACTCCGA | TACCTAATCA | TGATTAGGTA | TCAAGATAGT | ACTATCTTGA | ||
| A03 | GCGTTGGAGC | CCGGCGGCGA | TCGCCGCCGG | GGCCGTCAAC | GTTGACGGCC | E03 | GCTAACTTGC | AAGTAATATT | AATATTACTT | TAATTGACCT | AGGTCAATTA | ||
| A04 | GATCTTACGC | GGCTTGCCAA | TTGGCAAGCC | CCTAGACGAG | CTCGTCTAGG | E04 | AATCCATCTT | AGCTAAGAAT | ATTCTTAGCT | CAGCCGGCTT | AAGCCGGCTG | ||
| A05 | CTGATGGTCA | CCGCTAGCTG | CAGCTAGCGG | TCGTTAGAGC | GCTCTAACGA | E05 | GGCTGAGCTC | GTCGAGGTAT | ATACCTCGAC | AGAACCGGAG | CTCCGGTTCT | ||
| A06 | CCGAGAATCC | CTTATCCTAC | GTAGGATAAG | CGTTCTATCA | TGATAGAACG | E06 | CCGATTCCTG | TTATTAGTAG | CTACTAATAA | GAGATGCATG | CATGCATCTC | ||
| A07 | GCCGCAACGA | TGAGCTACTT | AAGTAGCTCA | CGGAATCTAA | TTAGATTCCG | E07 | ACCGCCAACC | TGCGAAGATC | GATCTTCGCA | GATTACCGGA | TCCGGTAATC | ||
| A08 | TGAGTCTGGC | TCAGGACTTA | TAAGTCCTGA | ATGACTGATC | GATCAGTCAT | E08 | TGGCCTGAAG | AACTACGGCT | AGCCGTAGTT | TCGTAACGGT | ACCGTTACGA | ||
| A09 | TGCGGACCTA | CCGCAGCCGC | GCGGCTGCGG | TCAATATCGA | TCGATATTGA | E09 | AACCTCATTC | AACGGAACGC | GCGTTCCGTT | TGGCGACGGA | TCCGTCGCCA | ||
| A10 | ACCTCGTTGA | TGCGCCTGGT | ACCAGGCGCA | GTAGACCTGG | CCAGGTCTAC | E10 | ATAAGGAGCA | GATGCTACGA | TCGTAGCATC | AGTCATAGCC | GGCTATGACT | ||
| A11 | ACGGAGGCGG | AATCATACGG | CCGTATGATT | TTATGACCAA | TTGGTCATAA | E11 | CGAACGCCGG | ATCTGCCAAT | ATTGGCAGAT | GTCAAGTCCA | TGGACTTGAC | ||
| A12 | TAGATCTACT | CGCCAATCAA | TTGATTGGCG | TTGGTCCGTT | AACGGACCAA | E12 | GGTATGCTTG | ATCGTATCAA | TTGATACGAT | ATTCGGAAGT | ACTTCCGAAT | ||
| B01 | AATTAAGACT | CAAGGCTTAG | CTAAGCCTTG | GGTACGTTAA | TTAACGTACC | F01 | AACCTGCGTA | AACGCCTCTA | TAGAGGCGTT | GTCGGTAGTT | AACTACCGAC | ||
| B02 | CCATTGCGTT | GCGCTCGACG | CGTCGAGCGC | CAATGAGTCC | GGACTCATTG | F02 | GGCAGACGCC | ACGGCAACCA | TGGTTGCCGT | AGGACGGACG | CGTCCGTCCT | ||
| B03 | TTATTCATTC | TCCAGCAATA | TATTGCTGGA | GATGCAGTTC | GAACTGCATC | F03 | TAGCCGTCAT | CAGGCTAAGA | TCTTAGCCTG | CTCCTGGACC | GGTCCAGGAG | ||
| B04 | ATCTCCGAAC | CATGAGAACT | AGTTCTCATG | CCATCGTTCC | GGAACGATGG | F04 | CCTGGAAGAG | CGCAATATCA | TGATATTGCG | TAGCCTCGTT | AACGAGGCTA | ||
| B05 | TTGACTTCAG | AACGTAATCT | AGATTACGTT | TTGAGAGAGT | ACTCTCTCAA | F05 | GGAGGTTCTA | TTCGATAACC | GGTTATCGAA | GGTTGAACGT | ACGTTCAACC | ||
| B06 | GGCAGGTATT | ATTCTCCTCT | AGAGGAGAAT | ACTGAGCGAC | GTCGCTCAGT | F06 | CTAGTAGTCT | AACCTCAAGA | TCTTGAGGTT | AGGTCCTCGT | ACGAGGACCT | ||
| B07 | AGAGCTATAA | TCTGCGCGTT | AACGCGCAGA | TGAGGAATCA | TGATTCCTCA | F07 | ATCATCAACG | CAGGCGCCAT | ATGGCGCCTG | GGAAGTTATA | TATAACTTCC | ||
| B08 | CTAAGAGAAG | GCTCATATGC | GCATATGAGC | CCTCCGACGG | CCGTCGGAGG | F08 | ACGCGAGATT | AACTATTATA | TATAATAGTT | TGGTAATCCT | AGGATTACCA | ||
| B09 | ACTCAATAGG | AGCGGTAACG | CGTTACCGCT | CATTGACGCT | AGCGTCAATG | F09 | GAAGAGGCAT | AAGTTACCTA | TAGGTAACTT | AAGCTAGGTT | AACCTAGCTT | ||
| B10 | CTTGCGCCGC | AATGAATAGT | ACTATTCATT | TCGTCCTTCG | CGAAGGACGA | F10 | GGTATCCGCC | CGGCAGAGGA | TCCTCTGCCG | TCCGCGGACT | AGTCCGCGGA | ||
| B11 | AATCGTAGCG | CCGTATCTGG | CCAGATACGG | TGATACTCAA | TTGAGTATCA | F11 | AACTAGGCGC | GCCTCAATAA | TTATTGAGGC | TGCGGATAGT | ACTATCCGCA | ||
| B12 | GGTACTGCCT | CCTTAGTCTG | CAGACTAAGG | TTCTACCTCA | TGAGGTAGAA | F12 | TCGCTAAGCA | TTAACGCCGT | ACGGCGTTAA | TGGCAGCTCG | CGAGCTGCCA | ||
| C01 | TAGAATTAAC | ACCTAGTTAG | CTAACTAGGT | TCGTCGGAAC | GTTCCGACGA | G01 | TATATACTAA | CATACGATGC | GCATCGTATG | TGCTACGGTC | GACCGTAGCA | ||
| C02 | GCCATTCTCC | ATAGGAGTAC | GTACTCCTAT | ATCGAGATGA | TCATCTCGAT | G02 | ACTTGCTAGA | AAGCTGACCT | AGGTCAGCTT | GCGCAATGAC | GTCATTGCGC | ||
| C03 | TGCCGGCAGA | CTACGACGAG | CTCGTCGTAG | TAGACTAGTC | GACTAGTCTA | G03 | AACCATTGGA | GAGTCCTTAT | ATAAGGACTC | CTTAATCTTG | CAAGATTAAG | ||
| C04 | TTACCGAGGC | AGTCGAGTTC | GAACTCGACT | GTCGAAGCAG | CTGCTTCGAC | G04 | TCGCGGTTGG | CCTACGGCAA | TTGCCGTAGG | GGAGTTGCGT | ACGCAACTCC | ||
| C05 | ATCATATTAG | TGGTCCAGTC | GACTGGACCA | AGGCGCTAGG | CCTAGCGCCT | G05 | CGTAGTTACC | AATATTCGAA | TTCGAATATT | ACTCGTATCA | TGATACGAGT | ||
| C06 | TGGTCAGCCA | ATCTAAGCAA | TTGCTTAGAT | AGATGCAACT | AGTTGCATCT | G06 | TCCAATCATC | TTCAAGAATC | GATTCTTGAA | GGTAATAATG | CATTATTACC | ||
| C07 | ACTATGCAAT | CGAATTCGTT | AACGAATTCG | AAGCCTACGA | TCGTAGGCTT | G07 | AATCGATAAT | ATGCTCGCAA | TTGCGAGCAT | TCCTTATAGA | TCTATAAGGA | ||
| C08 | CGACGCGACT | CAGCGATAGA | TCTATCGCTG | GTAGGCAATT | AATTGCCTAC | G08 | CCATTATCTA | GGAGTAAGCC | GGCTTACTCC | CCGACTCCAA | TTGGAGTCGG | ||
| C09 | GATACGGAAC | GGTCGCTATG | CATAGCGACC | TGCCAGTTGC | GCAACTGGCA | G09 | TCAACGTAAG | TTATCGTATT | AATACGATAA | GCCAAGCTTG | CAAGCTTGGC | ||
| C10 | TTATCCGGAT | ATCCGTTAGC | GCTAACGGAT | CTTAGGTATC | GATACCTAAG | G10 | TCTAATAGTA | AAGTCTAATA | TATTAGACTT | CATATCCTAT | ATAGGATATG | ||
| C11 | TAGAGTAATA | TCGCAATTAG | CTAATTGCGA | GAGACCTACC | GGTAGGTCTC | G11 | AACCGCTGGT | CGGCTTACTA | TAGTAAGCCG | ACCTACGCCA | TGGCGTAGGT | ||
| C12 | GCAGGTCCGT | GGCTGGCTAG | CTAGCCAGCC | ATTGACCGAG | CTCGGTCAAT | G12 | GATCGCTTCT | GATATGGTCT | AGACCATATC | GGAATTCAGT | ACTGAATTCC | ||
| D01 | TCGGCCTTAC | ACGGTCTTGC | GCAAGACCGT | GGAGGCGGCG | CGCCGCCTCC | H01 | CTAACTAGAT | TAGTCGTCCA | TGGACGACTA | TGGCGTAGAA | TTCTACGCCA | ||
| D02 | AGAACGTCTC | GCTCCATTCG | CGAATGGAGC | CCAGTACTTG | CAAGTACTGG | H02 | GCTGGAACTT | TAGCTGCTAC | GTAGCAGCTA | ATTGCGGCCA | TGGCCGCAAT | ||
| D03 | CCAGTTCCAA | ACGATAAGCG | CGCTTATCGT | GGTCTCGCCG | CGGCGAGACC | H03 | AGGTTAGTTC | CTCTTCAAGC | GCTTGAAGAG | TTCAGCTTGG | CCAAGCTGAA | ||
| D04 | GGCGTTAAGG | ACCATAGCGC | GCGCTATGGT | GGCGGAGGTC | GACCTCCGCC | H04 | CATTCGACGG | ATGAACGCGC | GCGCGTTCAT | CCATCTGGCA | TGCCAGATGG | ||
| D05 | ACTTAACCTT | CTCTTAGCGG | CCGCTAAGAG | TAGTTCTAGA | TCTAGAACTA | H05 | CATTCAATCA | GTCGACGGAA | TTCCGTCGAC | CTTATAAGTT | AACTTATAAG | ||
| D06 | CAACCGCTAA | TGATTCAACT | AGTTGAATCA | TTGGAGTTAG | CTAACTCCAA | H06 | CGGATTAGAA | ACTAATTGAG | CTCAATTAGT | GATTAGATGA | TCATCTAATC | ||
| D07 | GACCTTGATA | TATGGCCGCG | CGCGGCCATA | AGATCTTGGT | ACCAAGATCT | H07 | ATCGGCTATC | CTTGCATAAT | ATTATGCAAG | TATAGGATCT | AGATCCTATA | ||
| D08 | TCTGATACCA | AGAGGTCGCA | TGCGACCTCT | GTAATGATCG | CGATCATTAC | H08 | CCTTGATCGT | TCCTTACCAA | TTGGTAAGGA | AGCTTATAGG | CCTATAAGCT | ||
| D09 | GAAGATCGAG | AGGAGATTGA | TCAATCTCCT | CAGAGAGGTC | GACCTCTCTG | H09 | ACGAAGTCAA | TGCAGCCTAC | GTAGGCTGCA | GTCTGCAATC | GATTGCAGAC | ||
| D10 | AGGAGCGGTA | GGCTATATAG | CTATATAGCC | TTAATTAGCC | GGCTAATTAA | H10 | TTACCTCGAC | GGAGCTGAGG | CCTCAGCTCC | CGCCTCTTAT | ATAAGAGGCG | ||
| D11 | AAGAAGCTAG | TCGCGTACTT | AAGTACGCGA | CTCTAACTCG | CGAGTTAGAG | H11 | GGAGGATAGC | GCAGCGGACT | AGTCCGCTGC | GTTGGATCTT | AAGATCCAAC | ||
| D12 | TCCGGCCTCG | AATAATAATG | CATTATTATT | TACGATCATC | GATGATCGTA | H12 | GGCTCTCTAT | CATCGCGCTC | GAGCGCGATG | GCGATTGCAG | CTGCAATCGC |
Nuclei preparation
This is an exemplary protocol for one sample containing 1-5 x 10^6 starting cells. Perform all procedures on ice using chilled reagents and using a refrigerated swing bucket centrifuge.
Obtain a single-cell suspension using a dissociation protocol that preserves >95% cell viability. Eliminate any cell clump using a 40-100 μm cell strainer.
Extract and fix nuclei:
Centrifuge cells in a 15 mL conical at 150 g for 5 min, discard the supernatant, resuspend the pellet in 1 mL nuclei lysis buffer, and incubate for 5 min.
Centrifuge nuclei at 300 g for 3 min, discard the supernatant, re-suspend the pellet in 100 μL nuclei suspension buffer, add 10 mL 4% PFA, mix and incubate for 15 min.
Centrifuge fixed nuclei at 300 g for 5 min, discard the supernatant, and wash the pellet three times with 1 mL nuclei suspension buffer by repeating this step.
After the final wash, resuspend the pellet in 300 μL nuclei suspension buffer.
Count nuclei with a hemocytometer with YOYO-1 staining (Figure SP2.1), dilute them to 2,500 nuclei/μL, prepare single-use aliquots, and either proceed with the next section or snap-freeze nuclei in liquid nitrogen and store them at -80 °C for up to one month.
Figure SP2.1. Nuclear extraction QC
Exemplary picture of PFA-fixed nuclei stained with YOYO-1 visualized on a hemocytometer. Nuclei are round, not clumped, and there are minimal cell debris.
Reverse transcription
Clean the work area and instrumentation to prevent RNase contamination. Unless stated otherwise, keep samples on ice, perform centrifugations at 4 °C, use DNA LoBind 1.5 mL tubes and low-retention tips, and set the thermocycler heated lid at 105 °C to prevent evaporation.
Prepare fresh stop solution (40 mM EDTA and 1 mM spermidine; 600 µL/plate)
Thaw PFA-fixed nuclei in hand or at 37 °C thermal block, mix by flicking, then move to ice
Prepare a mix by combining nuclei suspension (2 µL, 5,000 nuclei, per well) and 10 mM dNTP (0.25 µL per well), and distribute 2.25 µL into each well of a 96-well LoBind plate
Add 1 µL of multiplexed RT primers to each well (step 1b of Reagent setup). Cap the plate and centrifuge it for 10 seconds at 100 g
Incubate the plate at 55 °C for 5 minutes and immediately place it on ice. After 3-5 minutes on ice, centrifuge the plate for 10 seconds at 100 g
Prepare the following RT mix, distribute 1.75 μL to each well without mixing, cap the plate, and centrifuge it for 10 seconds at 100 g
| A | B | C | |
| one well (ul) | one plate | ||
| 5X Superscript IV First-strand buffer | 1 ul | 105 ul | |
| 100mM DTT | 0.25 ul | 26.25 ul | |
| Superscript IV reverse transcriptase | 0.25 ul | 26.25 ul | |
| RNAseOUT Recombinant Ribonuclease Inhibitor | 0.25 ul | 26.25 ul |
Table SP2.2. iPS2-sci-seq RT mix
Reagent
Incubate the plate at 55 °C for 10 minutes, immediately place it on ice, and add 5 µL of stop solution into each well to stop the reaction. After 3-5 minutes on ice, centrifuge the plate for 10 seconds at 100 g.
Gently pool all wells into a FACS tube using wide bore tips, passing through a 40 µm cell strainer, and immediately proceed with the next procedure.
Nuclei sorting
Prepare collection plate(s) by adding 5 µL of elution buffer to each well one or more 96-well LoBind
plates, cover, and place on ice
Aliquot 50 µL of nuclei suspension (8 from Reverse transcription) in a separate tube to be used as
unstained control; add 3 µM DAPI to the remaining volume to stain nuclei
Setup gating for single nuclei (Fig. SP2.2), and sort 25 nuclei into each wells of collection plate
(assuming 96 indexes in the RT and an accepted rate of ~10% index collisions)
Note: use a cell sorter equipped with a 100 µm nozzle and minimize sample pressure, chill both
sample and collection plate, and mix the sample while sorting
Centrifuge the plates for 1 minute at 900 g, and either proceed with the next step or store plates at
-80 °C for up to one week
Figure SP2.2. Nuclei FACS Exemplary gating strategy for single nuclei sorting. Set up the gates based on the negative control (top) and sort the DAPI+ population. DAPI+ single nuclei are back-gated in red; debris are in light blue.
Second strand synthesis
Prepare reaction master mix, distribute 1 µL to each well, pipette mix, cap the plate, and centrifuge
it for 10 seconds at 100 g
| A | B | C | |
| one well (ul) | one plate | ||
| mRNA Second Strand Synthesis buffer (NEB) | 0.6ul | 67.5ul | |
| mRNA Second Strand Synthesis enzyme (NEB) | 0.3ul | 33.75ul | |
| mol bio ddH2O | 0.1ul | 11.25ul |
Table SP2.3. iPS2-sci-seq second strand synthesis mix
Carry out second strand synthesis at 16 °C for 180 minutes, and terminate the reaction by incubating
at 75 °C for 20 minutes (lid kept at 80 °C for both steps)
Centrifuge the plate for 10 seconds at 100 g and either immediately proceed with the next section
or store at 4 °C overnight
Tagmentation
Prepare the tagmentation mix with the Nextera TD buffer (5.75 µL/well) and the TDE1 enzyme (0.01 µL/well), distribute 5.76 µL of mix to each well of the sorting plate, pipette mix, cap the plate, and centrifuge it for 10 seconds at 100 g
Incubate the plate at 55 °C for 5 min, immediately place it on ice, add 12 µL of Zymo DNA binding buffer to each well, and incubate at room temperature for 5 min.
Purify tagmented dsDNA using 1.5X volumes of SPRIselect beads (36 µL/well), following the manufacturer’s instructions and eluting in 17 µL of elution buffer
Transfer 16 µL of eluted DNA to a new 96-well LoBind plate and either proceed with the next section or store at 4 °C overnight
Indexing PCR and NGS
Add 2 µL of 10 µM P5 primer and 2 µL of 10 µM multiplexed P7 primers (step 1e from Reagent setup) to each well, applying a unique combination of i5 and i7
Add 20 µL NEBNext High-Fidelity 2X PCR Master Mix to each well, pipette mix, cap the plate, and centrifuge it for 10 seconds at 100 g, and run the following PCR:
| cycles | temperature | timing | step | |
| 1 | 75°C | 3 min | ||
| 1 | 98°C | 30 sec | Initial Denaturation | |
| 18-22 cycles | 98°C | 10 sec | Denaturation | |
| 66°C | 30 sec | annealing | ||
| 72°C | 1 min | extension | ||
| 1 | 72°C | 5 min | final extension |
Table SP2.4. iPS2-sci-seq indexing PCR cycling conditions
Pool all samples of a 96-well plate in a 15 mL Falcon tube and concentrate on 4 columns of the Zymo DNA Clean & Concentrator kit, following the manufacturer’s instructions except without performing any washes and eluting in a total volume 25 µL of elution buffer
Further purify using 0.85X volumes of SPRIselect beads, according to the manufacturer’s instruc-
tions and eluting in 50 μL of elution buffer
Run 2 μL of a 1:10 dilution of the NGS library on a TapeStation High Sensitivity 5000 Screen Tape
assay, following the manual. Expect a main peak at ~350 bp (Figure SP2.3)
Perform NGS using NextSeq 1000/2000 and 50 cycles reagents: read 1 - 18 cycles; index 1 &
index 2 - 10 cycles each; read 2 - 52 cycles (>25,000 reads/nuclei)
Figure SP2.3. iPS2-sci-seq NGS library QC
Exemplary TapeStation QC of iPS2-sci-seq NGS library (1:10 dilution).
Protocol 3: iPS2-10X-seq library preparation
iPS2-10X-seq relies on microfluidics partitioning-based scRNA-seq from 10X Genomics. Despite the higher costs, the ease and reproducibility of this commercial method have led to its widespread adoption. The procedure integrates the manufacturer’s protocol to generate an additional NGS library containing UCI-BCs. This is sequenced in parallel to the standard gene expression library to enable in silico matching of UCI-BCs and single-cell transcriptomes through their shared unique cell barcodes. Of note, iPS2-10X-seq provides built in support for the cost-saving strategy of reasonably overloading the microfluidics chip, since cell doublets are readily identified by the co-expression of two different UCI-BCs and filtered out as part of the standard catcheR_10Xcatch pipeline. Thus, it is possible to load the manufacturer-recommended cell number for barcoded sample pools even without additional multiplexing barcodes.
Other cost-saving strategies include using cell multiplexing oligos (CMOs) or barcoded antibodies against housekeeping cell surface antigens to multiplex tet-untreated and tet-treated cells in a single microfluidics channel, enabling robust control over knockdown experimental design without performing two separate reactions. This Supplemental Protocol describes the following procedures:
1. Prepare the gene expression library
2. Prepare the shRNA barcode library
3. Library pooling and NGS
Prepare the gene expression library
Obtain a single-cell suspension using a dissociation protocol that preserves >95% cell viability. Avoid the use of DNase, and remove any EDTA with at least three washes.
OPTIONAL: Label multiple samples using TotalSeq-B hashtags and/or other antibodies with ADTs, following the manufacturer’s instructions (CG000149 Rev D).
OPTIONAL: Label multiple samples using cell multiplexing oligos (CMOs) in the 3’ CellPlex Kit from 10X Genomics, following the manufacturer’s instructions (CG000391 Rev B).
OPTIONAL: Sort live cells after staining with Fixable Viability Dye eFluor 780.
Note: this step increases the fraction of cells that deliver high-quality transcriptomes, and is particularly recommended for organoids that are difficult to dissociate. Avoid the use of other live/dead staining that can intercalate with DNA.
Centrifuge cells/cell pools at 150 g for 5 min at 4°C, and resuspend in ice cold DPBS 1% BSA aiming for 1 x 10^6^ cells/mL.
Manually count cells, and calculate the cell volume needed to load the target cell number per channel of Chromium Next GEM Chip G.
Note: since iPS2-seq cells express shRNA barcodes, chips can be reasonably over-loaded even without any CMO or hashtags (~49,500 cells/reaction at a ~24% doublet rate).
Follow the 10X Genomics protocols to obtain and quality control the gene expression library (GEX) and, when appropriate, the CMO library and/or the ADT library.
Save a portion of the pre-amplified full-length cDNA for the next procedure.
Prepare the shRNA barcode library
Perform a two-step PCR to enrich the UCI-BC region in the 3’ UTR of the OPTtetR cDNA.
Perform the first PCR according to the calculations and cycling conditions below:
| A | B | |
| Reagent | Volume (ul) | |
| NEBNex High-Fidelity 2X PCR Master Mix | 10 ul | |
| iPS2-10X-seq_truseq | 1 ul | |
| iPS2-10X-seq_inner | 1 ul | |
| cDNA 30-60ng/uL | variable | |
| H2O (bring to volume) | to 10ul | |
| Total volume | 20 ul |
Table SP3.1. iPS2-10X-seq first PCR mix
| A | B | C | D | |
| Steps | Temperature | Time | Cycles | |
| Initial denaturation | 98°C | 30 sec | 1 | |
| Denaturation | 98°C | 10 sec | 15 | |
| Annealing | 60°C | 20 sec | ||
| Extension | 72°C | 20 sec | ||
| Final extension | 72°C | 30 sec | 1 |
Table SP3.2. Cycling conditions for iPS2-10X-seq first PCR
Note: When performing iPS2-multi-seq, adjust this step by increasing the input material to 200 ng and reducing the cycle amplification to 11 cycles. Adaptations of the current protocol were essential to obtain clean UCI-BC library profiles.
Purify the PCR with 1.8X volumes of SPRIselect beads, following the manufacturer’s instructions and eluting in 20 µL of elution buffer.
Run and quantify 2 µL of undiluted PCR on a TapeStation High Sensitivity 5000 Screen Tape assay, following the manual. Expect a main peak at ~285 bp (Figure SP3.1A).
Dilute the first PCR to 0.1 ng/µL and use it as a template for the second PCR according to the calculations and cycling conditions below:
| A | B | |
| Reagent | Volume (µL) | |
| NEBNext High-Fidelity 2X PCR Master Mix | 10 | |
| Dual Index Kit TT Set A primers | 4 | |
| First PCR (0.1 ng/uL) | variable | |
| Nuclease-free water | to 20ul | |
| Total volume | 20 |
Table SP3.3. iPS2-10X-seq second PCR mix
| A | B | C | D | |
| Step | Temperature | Timing | Cycles | |
| Initial denaturation | 98 °C | 30 sec | 1 | |
| Denaturation | 98 °C | 10 sec | 10 | |
| annealing | 58 °C | 20 sec | ||
| extension | 72 °C | 20 sec | ||
| final extension | 72 °C | 30 sec | 1 |
Table SP2.7. Cycling conditions for iPS2-10X-seq second PCR
Note: use a different index from the one utilized to prepare the GEX, CMO, and/or library, to allow subsequent demultiplexing of the UCI-BC library.
Purify and perform a double-sided size selection using 0.7X 26 0.9X volumes of SPRIselect beads, following the manufacturer’s instructions and eluting in 20 µL of elution buffer.
Run 2 µL of undiluted PCR on a TapeStation High Sensitivity 5000 Screen Tape assay, following the manual. Expect a main peak at ~385 bp (Figure SP3.1B).
Figure SP3.1. Exemplary TapeStation QC of iPS2-10X-seq first (A) and second (B) PCR.
Library pooling and NGS
Quantify the GEX and, when appropriate, CMO and/or ADT libraries (step 3 of Prepare the gene expression library), and the UCI-BC library (step 11 of Prepare the shRNA barcode library) with a Qubit dsDNA High Sensitivity kit.
Pool GEX, CMO, ADT and UCI-BC libraries at a relative molar ratio of 60:10:10:3.
Perform NGS using NextSeq 1000/2000 and 100 cycles reagents, with the following settings: read 1 - 28 cycles; index 1 - 10 cycles; index 2 - 10 cycles each; read 2 - 90 cycles ( 3e30,000 reads/target cells).
Protocol 4: Data processing and analysis with catcheR
iPS2-seq design and analysis with catcheR
This Supplemental Protocol describes the following procedures:
- Overview
- Installation
- Oligonucleotides design
- Pooled cloning step 1 QC
- Pooled cloning step 2 and hiPSC genome editing QC
- iPS2-10X-seq perturbation deconvolution
- iPS2-sci-seq perturbation deconvolution
- Barcode reassignment
- Perturbation effect analysis
Overview
catcheR is a comprehensive bioinformatic package for designing and analyzing iPS2-seq experiments. For the complete documentation, see https://marialuisaratto.github.io/catcheRdocs/
It comprises the following functions:
Figure SP4.1. Overview of catcheR
Logical relationship and main inputs/outputs for the catcheR analytical pipeline to assign perturbation in iPS2-seq experiments. Functions used for oligo design, plasmid/hiPSC QC, and perturbation effect analyses are not illustrated.
catcheR_design, which designs oligonucleotides for Supplemental Protocol 1 - Design shRNA oligonucleotides, facilitating shRNA library cloning.
catcheR_step1QC, which analyzes the results of Supplemental Protocol 1 - Intermediate plasmid pool QC, assessing pooled cloning step 1 plasmids for barcode swaps.
catcheR_step2QC, which analyzes the results of Supplemental Protocol 1 - Final plasmid pool QC or hiPSC pool QC, assessing pooled cloning step 2 or genome-edited hiPSC pools for shRNA representation.
catcheR_scicount, which analyzes 2-level indexing sci-RNA-seq data, facilitating the generation of gene expression matrix for iPS2-sci-seq experiments.
catcheR_scicatch, which assigns shRNA perturbations to single nuclei transcriptomes obtained by Supplemental Protocol 2, enabling the primary analysis of iPS2-sci-seq.
catcheR_10Xcatch, which assigns shRNA perturbations to single cell transcriptomes obtained by Supplemental Protocol 3, enabling the primary analysis of iPS2-10X-seq.
catcheR_scicatchQC and catcheR_10XcatchQC, which use the outputs of catcheR_scicatch and catcheR_10Xcatch, respectively, to fine-tune shRNA assignment thresholds.
catcheR_filtercatch, which leverages on the output of catcheR_scicatchQC and catcheR_10XcatchQC to filter single nuclei/cell transcriptomes expressing a single shRNA.
catcheR_sortcatch, which quality controls the cell-by-gene matrix based on the results of catcheR_step1QC, reassigning hPSC clones with barcode swaps to the correct shRNA.
catcheR_scinoatch and catcheR_10Xnoatch, which identify cells expressing no shRNA in iPS2-sci-seq and iPS2-10X-seq experiments, respectively, adding them to the cell-by-gene matrix to be used as additional controls.
catcheR_load, which loads gene expression matrices annotated with shRNA perturbations into a Monocle object, preparing the dataset for downstream analysis.
catcheR_pseudotime, which analyzes the effects of shRNA perturbations on pseudotime dynamics, highlighting shifts along differentiation trajectories (e.g., Figures 2J, 2K).
catcheR_modules, which assesses perturbation-induced changes in gene module expression, such as coordinated activation or repression of functional programs (e.g., Figures 5H–5J).
catcheR_enrichment, which quantifies differences in perturbation representation across experimental samples (e.g., Figures 5F, 5G) or cell clusters (e.g., Figure 5E).
Installation
catcheR is available at https://github.com/alessandro-bertero/catcheR. The GitHub repository folder "scripts" contains all the bash and R scripts that can be run independently. However, in order to ensure reproducible analyses, we strongly recommend to install catcheR package from GitHub, since its functions run all the analysis inside Docker containers.
Install Docker engine, following the instructions at https://docs.docker.com/engine/install/
Install catcheR
In R (≥ 3.0.2) install devtools, if not already present
install.packages("devtools")
Install catcheR from GitHub
install_github("alessandro-bertero/catcheR")
Install rundocker from GitHub
install_github("Reproducible-Bioinformatics/rundocker")
Load catcheR and rundocker in your R environment
library(catcheR)
library(rrundocker)
Oligonucleotides design
In a new working folder, prepare the following files:
- A comma-separated values (CSV) file with three columns listing: (1) the forward oligos from the TRC shRNA library; (2) the corresponding barcodes (BC); (3) the shRNA names. Below is an example for the shRNA described in Figure SP1.2:
CCGGCAAACTCCTCTGCGATTCTCGGAACATCCAGACAGAGACTGTTTTTG,CAGTTCCA,SMAD2.1
- (Optional) - A txt file with a newline-separated list of 5’-3’ restriction sites, or other sequences, to be avoided in the shRNAs. By default these are SalI, SwaI and AscI:
GTCGAC
ATTTAAAT
GGCGCGCC
Run catcheR_design:
catcheR_design(
group=c("docker","sudo"),
folder,
sequences,
gibson.five = "AGTTCCTCTACTAGTGATAGAGATCCC",
gibson.three = "GTAGCTCGGTGATCAGC",
fixed = "GTCGACATTTAAATGGCGCGCC",
restriction.sites = NULL)
catcheR_design arguments:
- group: string with two options: sudo or docker, depending on the user group. For a detailed explanation of Docker user groups, see this page
- folder: string with the working folder path
- sequences: string with the CSV file name from step 1a
- gibson.five: (optional) string with the 5’ Gibson homology
- gibson.three: (optional) string with the 3’ Gibson homology
- fixed: (optional) character string with the multicloning site
- restriction.sites: (optional) string with the txt file name from Step 25
Example usage:
catcheR_design(
group = "docker",
folder = "path/to/folder",
sequences = "filename.csv",
restriction.sites = "filename.txt")
Pooled cloning step 1 QC
This step complements Supplemental Protocol 1 - Intermediate plasmid pool QC.
In a new working folder, prepare the following files:
- (a) Fastq/fq or fastq.gz files with demultiplexed read 1 from the NGS run
- (b) A CSV file with the shRNA names and their full sequences
SMAD2.1,GCAGATCCTCTCTGCTGGATTCTCGGAACATCCAGACAGAGACTG
- (c) (Optional) - A txt file with a newline-separated list of clones of interest as "BC_UCI"
CAAGACCC_CATCGT
Run catcheR_step1QC:
catcheR_step1QC(
group=c("docker","sudo"),
folder,
fastq.read1,
DIs = 100,
ratio = 10,
plot.threshold = 2000,
clones = NULL)
catcheR_step1QC arguments:
- (a) group: string with two options: sudo or docker, depending on the user group (info)
- (b) folder: string with the working folder path
- (c) fastq.read1: string with the read 1 filename from step 1a
- (d) DIs: integer of the minimum number of diversity indexes (DIs, pseudo-unique reads) of the most represented shRNA matching to a given UCI-BC; in combination with ”ratio”, it selects UCI-BCs for which it is possible to reliably assign an shRNA
- (e) ratio: integer of the minimum ratio between the number of DIs of the most represented and second most represented shRNAs matching to a given UCI-BC
- (f) plot.threshold: integer of the minimum number of DIs per UCI-BC for output 2b
- (g) clones: (optional) a string with the txt file from step 1c
Example usage:
catcheR_step1QC (
group = "docker",
folder = "path/to/folder",
fastq.read1 = " filename .fq",
clones = " filename .txt ")
catcheR_step1QC key outputs:
- (a) ”reliable_clones_swaps.csv”, which lists UCI-BCs with reliable evidence of shRNA-barcode swap, and is used as input for Barcode reassignment
- (b) Bar chart of the number of DIs for each clone above ”plot.threshold”
- (c) Bar chart of the number of DIs associated to each BCs, shRNAs, and reliable swaps
- (d) (If ”clones” argument is provided) - CSV files and bar charts with the number of DIs for each shRNA matching to each clone of interest (Figure SP4.2)
Figure SP4.2. Example of quantification of DIs for bacterial clones carrying the expected
shRNA (A) or with robust evidence of a barcode swap that can be reassigned (B; Figure 1B).
These graphs allow visual evaluation of appropriate thresholds for ”DI” and ”ratio”.
Pooled cloning step 2 and hiPSC genome editing QC
In a new working folder, prepare the following files:
- (a) Fastq/fq or fastq.gz files with demultiplexed read 1 from the NGS run
- (b) "rc_barcodes_genes.csv", a CSV file with two columns: (1) the shRNA BCs; (2) the matching shRNA names in the format "GENE.shRNAID"
- (c) (Optional) - A txt file with clones of interest (step 1c of Pooled cloning step 1 QC)
Run catcheR_step2QC:
catcheR_step2QC(
group=c("docker","sudo"),
folder,
fastq.read1,
DIs = 1000,
clones = NULL)
catcheR_step2QC arguments:
- (a) group: string with two options: sudo or docker, depending on the user group
- (b) folder: string with the working folder path
- (c) fastq.read1: string with the read 1 file from step 1c
- (d) DIs: integer of the minimum number of DIs for a given UCI-BC; it selects reliably-measured UCI-BCs
- (e) clones: (optional) a string with the txt file from step 1c
Example usage:
catcheR_step2QC(
group = "docker",
folder = "path/to/working/folder",
fastq.read1 = "filename.fq",
clones = "filename.txt")
catcheR_step2QC key outputs:
- (a) Pie charts of DIs per shRNA and per gene targets associated to each clone
- (b) Text file listing all clones above the DI threshold
- (c) Bar chart of the number of DIs for each clone above the DI threshold (Figure SP4.3A)
- (d) Frequency histogram of the percentage of clones above the DI threshold for each shRNA (Figure SP4.3B)
A
B
Figure SP4.3. Example of quantification of DIs for bacterial clones represented by more than 1,000 DIs (A), and of the frequency of bacterial clones for each shRNA. These graphs confirm whether shRNAs are normally distributed in the final plasmid pool.
Obtain iPS2-10X-seq, iPS2-CITE-seq, or iPS2-multi-seq count matrice
Download and install cellranger (for iPS2-10X-seq and iPS2-CITE-seq) or cellranger-arc (for iPS2-multi-seq). Alternatively, Docker containers for cellranger v7, cellranger v9 (recommended for iPS2-CITE-seq), or cellranger-arc can be pulled from our Docker repository.
For iPS2-10X-seq and iPS2-CITE-seq, demultiplex Illumina base call (BCL) files using cellranger mkfastq, following the standard 10X Genomics user guide. In the sample sheet CSV, include the index sequences used in Supplemental Protocol 3 for the GEX and UCI-BC libraries, and, when applicable, CMO and/or ADT libraries.
For iPS2-multi-seq, use cellranger-arc mkfastq to demultiplex GEX + ATAC dual-index libraries, following the Cell Ranger ARC user guide. Ensure the sample sheet format is compatible with dual-modality sequencing runs, and include the index sequences used for both gene expression and chromatin accessibility libraries.
Run FastQC to assess the quality of the FASTQ files for each library type.
Generate cell-by-gene count matrices using cellranger count (for single-sample experiments) or cellranger multi (for multiplexed experiments, including iPS2-CITE-seq). For iPS2-multi-seq, use cellranger-arc count to obtain both gene expression and chromatin accessibility matrices.
Note: In multiplexed experiments (e.g., using CMO or ADT barcodes in iPS2-CITE-seq), count matrices from individual samples can be aggregated using cellranger aggr to produce a unified dataset. This enables joint analysis in a single run of catcheR_10Xcatch by specifying the number of samples via the samples argument.
Use cellranger mat2csv to convert the sparse matrix output into a dense CSV format for compatibility with downstream tools. For iPS2-multi-seq, run cellranger-arc mat2csv separately for GEX and ATAC outputs, if needed.
After obtaining the count matrix, proceed to iPS2-10X-seq perturbation deconvolution.
Obtain iPS2-sci-seq count matrices
catcheR_scicount is a wrapper of the "bbi-sci" pipeline developed by the Brotman Baty Institute for Precision Medicine, which was implemented and dockerized in catcheR to be used on any operating system.
Demultiplex Illumina base calls to fastq files
- (a) Create a "SampleSheet.csv" file with as many sample rows as the PCR wells from Indexing PCR and NGS of Supplemental Protocol 2, where "Sample_ID" is the well identified with format [A-H]01-[12], and index and index2 are the i7 and i5 indexes used in the corresponding row and column (refer to Table SP2.1)
- (b) Run Illumina bcl2fastq following Illumina’s manual
- (c) Run fastQC to confirm the quality of the fastq files
In a new working folder:
- (a) Create the subfolder "fastq", and copy all the demultiplexed fastq.gz files. Ensure all file names begin with the well coordinate (e.g. A01) followed by an underscore
- (b) Create a tab separated txt file called "sci-RNA-seq-8.RT.oligos" with the association between RT wells and RT barcode sequences (refer to Table SP2.1)
Run catcheR_scicount:
catcheR_scicount(
group=c("docker","sudo"),
folder,
sample.name,
UMI.cutoff)
catcheR_scicount arguments:
- (b) folder: string with the working folder path
- (c) sample.name: string with the name of the experiment
- (d) UMI.cutoff: integer of the minimum number of UMI per nucleus needed to consider the single cell transcriptome valid
Example usage:
catcheR_scicount(
group = "docker",
folder = "path/to/file",
sample.name = "experiment",
UMI.cutoff = 500)
catcheR_scicount outputs, found in the final-output folder: knee-plot of UMI per cell (Figure S1H), statistics files, sparse cell-by-gene count matrix, dense gene matrix "exp_mat.csv" and "exp_mat_no0.csv" (filtered to exclude genes with 0 counts in all cells), also in Rdata format.
After obtaining the count matrix, proceed to iPS2-sci-seq perturbation deconvolution.
iPS2-10X-seq perturbation deconvolution
shRNA perturbations can be assigned to single cells using catcheR_10Xcatch, which identifies NGS reads containing UCI-BCs, matches them to the corresponding transcriptome based on their shared cellular barcodes, applies filters for background noise, and selects cells with robust evidence of a single integration. UCI-BC filtering involves: (1) filtering out UCI-BCs supported by less than a certain number of UMIs in a given transcriptome, to account for PCR and sequencing artifacts; (2) filtering out UCI-BCs that represent less than a certain fraction of all UMIs for UCI-BCs in a given transcriptome, to reduce noise arising from free mRNA; and (3) filtering out UCI-BCs whose UMI fraction in a given transcriptome is not several fold greater than the second most common UCI-BC, to eliminate cells that contain multiple UCI-BCs when the second most common one falls just below one of the first two thresholds. The resulting list of bona fide shRNA integrations per cell is then used to filter those expressing a single shRNA. This section outlines a typical analysis workflow for iPS2-10X-seq, iPS2-CITE-seq or iPS2-multi-seq experiments, which follow analogous preprocessing and count matrix structures based on 10X Genomics technology. The pipeline begins with catcheR_10Xcatch, which performs a full analysis using automatic thresholding. If needed, these thresholds can be subsequently refined with catcheR_10XcatchQC and applied for re-filtering using catcheR_filtercatch. Finally, catcheR_10Xnocatch allows incorporation of cells that do not express any shRNA as additional negative controls in the final annotated cell-by-gene matrix.
In a new working folder:
Copy the fastq/fq or fastq.gz files with demultiplexed read 1 and read 2 of the UCI-BC library (step 3 of Obtain iPS2-10X-seq, iPS2-CITE-seq, or iPS2-multi-seq count matrices)
Copy the cell-by-gene count matrix in CSV format (step 6 of Obtain iPS2-10X-seq, iPS2-CITE-seq, or iPS2-multi-seq count matrices)
Create a CSV file named “rc_barcodes_genes.csv” with two columns: (1) the shRNA BCs; (2) the matching shRNA names in the format “GENE.shRNAID”
Run catcheR_10Xcatch to execute the complete analysis with automatic thresholding:
catcheR_10Xcatch(
group=c("docker","sudo"),
folder,
fastq.read1, fastq.read2,
expression.matrix,
reference = "GGGCCGCTCATCTGGGGAGCCG",
UCI.length = 6,
threads = 2,
percentage = 15,
mode = "bimodal",
ratio = 5,
samples = 1,
x = 100, y = 400)
Example usage:
catcheR_10Xcatch(group = "docker",
folder = "path/to/folder",
fastq.read1 = "R1.fastq.gz", fastq.read2 = "R2.fastq.gz",
expression.matrix = "filename.csv",
threads = 12)
catcheR_10Xcatch arguments:
- (b) folder: string with the working folder path
- (c) fastq.read1: string with the filename of read 1 fastq/fq or fastq.gz (step 1a)
- (d) fastq.read2: string with the filename of read 2 fastq/fq or fastq.gz (step 1a)
- (e) expression.matrix: string with the filename of the count matrix file CSV (step 1b)
- (f) reference: string with the reverse complement of the sequence before the shRNA-BC at the start of read 2 (default is the 3’ end of the OPTtetR cDNA, optional argument; Figure S2F)
- (g) UCI.length: integer of the UCI length (default is 6, optional argument)
- (h) threads: integer of the threads to be used for parallelization (default is 2, optional argument)
- (i) percentage: integer of the percentage of UMIs supporting a given UCI over the total UMIs supporting all UCIs in a given cell; it is used as threshold to consider the UCI valid (Figure SP4.4D); the recommended default is 15 (optional argument)
- (j) mode: string with two options: bimodal or noise, defining the mode for automatic thresholding the minimum number of UMIs per UCI to consider the UCI valid (optional argument)
- ”bimodal” (default) sets the threshold at the valley of the bimodal UMIxUCI distribution (Figure SP4.4C)
- ”noise” sets the threshold at 1.35 * the number of UCIs supported by a single UMI; used if the UMIxUCI distribution is not bimodal
- (k) ratio: the ratio between the number of UMIs supporting the most represented UCI and the number of UMIs supporting the second most represented UCI. The parameter is needed to identify the cell as a single integration cell. The default is 5 (optional argument)
- (l) samples: the number of samples present in the same experiment (cells from different experiments will have the corresponding number after the cell ID in the gene expression matrix, so that multiple reactions of the same experiment can be unified). The default is 1 (optional argument)
- (m) x: an integer indicating the upper limit on the x-axis of the cropped version of the plot ”UMIxUCI”. Default is 100 (optional argument)
- (n) y: an integer indicating the upper limit on the y axis of the cropped version of the plot ”UMIxUCI”. Default is 400 (optional argument)
catcheR_10Xcatch key outputs (will be stored inside the "Result" folder of each sample):
- (a) ”log.txt” with the starting number of reads
- (b) ”log2.txt” with the number of cells, UMIs, UCIs, and the calculated thresholds of UMI per UCI for both ”bimodal” and ”noise” (only one is chosen based on the ”mode” argument; the other one is provided for a potential reanalysis)
- (c) Bar charts of UMI counts per shRNA and target gene (Figure SP4.4A and Figure SP4.4B)
- (d) Frequency histograms of UCIs supported by a certain number of UMIs (UMIxUCI), to interpret the signal/noise of the experiment and possibly set a custom UMIxUCI threshold for subsequent reanalysis (Figure SP4.4C) Note: here and below, copies of the same UCI expressed by different cells are plotted and analyzed separately
- (e) Frequency histogram of UCIs supported by a certain fraction of UMIs over the total number of UMIs supporting all UCIs in a given cell (UMIpercentagexUCI), to further assess signal/noise and possibly adjust the default threshold (Figure SP4.4D
- (f) Dot plots that combine the UMIpercentagexUCI and UMIxUCI data on the x and y axes, respectively, with each dot representing one or more UCI (quantified by the dot size). Dot colors indicate either the number of valid UCIs (i.e., shRNAs) in the cell containing a given UCI (Figure SP4.4E), or whether said UCI is the only valid one in the relevant cell and is thus assigned to such cell (Figure SP4.4F)
- (g) ”log_part3.txt” with how many cells were assigned to a single shRNA or were filtered out due to zero or multiple shRNAs
- (h) ”silencing_matrix.csv” is the key output used for the secondary analyses: the cell-by-gene count matrix provided as input, filtered and annotated with the shRNA encoded in each cell. It is also provided in RDS format for easy loading into R. Cell names are modified as follows:
cellID_UMIxUCI_BC_GENE_UCI
Where:
- cellID is the original cell name (i.e., the 16 bp of the 10X RT barcode)
- UMIsxUCI is the number of UMI associated with that perturbation
- BC is the shRNA barcode
- GENE is the shRNA target
- UCI is the Unique Clonal Identifier (shared by all cells originating from the same hPSC clone)
Example of a cell name after the analysis:
TTCTAACCACAGTCGC_180_CGTGATGC_NKX2 .5 _ACAGTG
The annotated matrix can be directly used with the functions implemented in catcheR (see Perturbation effect analysis) to perform a range of secondary analyses described in the manuscript. Additional scripts are available at our GitHub repository for customized or extended analyses, most of which are now incorporated into catcheR.
A
B
C
D
E
F
Figure SP4.4. Exemplary key outputs of a catcheR_10Xcatch analysis: ”bar-
code_distribution.pdf” (A); ”gene_distribution.pdf” (B); ”UMIxUCI_400_100.pdf”
with the automatic UMIxUCI threshold based on ”bimodal” (C); ”per-
centage_of_UMIxUCI_dist.pdf” with the default 15% ”percentage” thresh-
old (D); ”2D_percentage_of_UMIxUCI_UMI_count_trueorfalse.pdf” (E); and
”2D_percentage_of_UMIxUCI_UMI_ValidCells.pdf” (F).
OPTIONAL: fine tune iPS2-10X-seq perturbation assignment
In the same working folder used to run catcheR_10Xcatch, run catcheR_10XcatchQC, which uses the txt file outputs to regenerate the quality control plots described in Figure SP4.4 and suggest new thresholds to enable subsequent filtering.
catcheR_10XcatchQC(
group=c("docker","sudo"),
folder,
reference = "GCGCGCTTCATCTCGGGGAGCCG",
mode = "bimodal",
sample = 1
x = 100, y = 400)
Example usage:
catcheR_10XcatchQC(
group = "docker",
folder = "path/to/working/folder",
mode = "noise")
The function is to be run separately for each sample (specified by argument sample with default 1). catcheR_10XcatchQC arguments and outputs are the same as catcheR_10Xcatch.
In the same working folder used to run catcheR_10Xcatch, run catcheR_filtercatch, which uses the output of catcheR_10Xcatch to filter cells based on custom thresholds.
catcheR_filtercatch(
group=c("docker","sudo"),
folder,
expression.matrix,
UMI.count,
percentage = 15,
ratio = 5,
sample = 1)
Example usage:
catcheR_filtercatch(group = "docker",
folder = "path/to/working/folder",
expression.matrix = "filename.csv",
UMI.count = 5, sample = 1)
The function is to be run separately for each sample (specified by argument sample with default 1). catcheR_filtercatch argument percentage is the minimum percentage of UMI for a given UCI over the total UMIs of UCIs in that cell, to consider the UCI valid. UMI.count is an integer of the custom UMI threshold. ratio is the minimum ratio between UMIs of top UCI and UMIs of 2nd top UCI. The other arguments and the outputs are the same as catcheR_10Xcatch.
OPTIONAL: identify cells expressing no shRNA in iPS2-10X-seq
In the same working folder used to run catcheR_10Xcatch, run catcheR_10Xnocatch:
catcheR_10Xnocatch(
group=c("docker","sudo"),
folder,
expression.matrix,
threshold,
sample = 1,
reference = "TAGCCGTTCACTCGGGGACGCG")
Example usage:
catcheR_10Xnocatch(group = "docker",
folder = "path/to/folder",
expression.matrix = "filename.csv",
threshold = 5)
catcheR_10Xnocatch argument threshold is an integer of the minimal UMI count for the "empty" reference (Figure S2F) needed to confidently identify it as having integrated said plasmid. All other arguments are the same as catcheR_10Xcatch.
catcheR_10Xnocatch output is an updated "silencing_matrix.csv", also available in RDS format, in which names of empty cells are modified as "cellID__?_empty_NA_empty".
OPTIONAL: merge multiple samples
In case fine-tuning is applied after catcheR_10Xcatch and multiple samples are present, run catcheR_merge to aggregate the results in a single matrix. This step is done automatically when running catcheR_10Xcatch.
catcheR_merge(
group=c("docker","sudo"),
folder,
samples = 2,
empty = T)
The argument "empty" determines whether the cells identified by catcheR_nocatch should be added to the matrix.
Example usage:
catcheR_merge(group = "docker",
folder = "/20tb/ratto/catcheR/test_CM5/",
samples = 4)
iPS2-sci-seq perturbation deconvolution
This section describes a typical analysis pipeline similar to the one described in the previous section, but leveraging on the catcheRscicatch, catcheRscicatchQC, catcheRfiltercatch, and catcheRscinoactch functions.
In a new working folder:
Create a subfolder called "fastq", and copy all the demultiplexed fastq.gz files (step 2 of Obtain iPS2-sci-seq count matrices). Ensure that all file names begin with the well coordinate (e.g. A01)
Copy the cell-by-gene expression matrix CSV file obtained with catcheR_scicount function (step 4 of Obtain iPS2-sci-seq count matrices)
Create a CSV file called "rcbarcodesgenes.csv" with two columns: (1) the shRNA BCs; (2) the matching shRNA names in the format "GENE.shRNAID"
Copy the txt file "sci-RNA-seq-8.RT.oligos" used by catcheR_scicount (step 3b of Obtain iPS2-sci-seq count matrices)
Run catcheR_scicatch to execute the complete analysis with automatic thresholding; the arguments are the same as for catcheR_10Xcatch (step 2 of iPS2-10X-seq perturbation deconvolution), except that no filenames are provided since fastq files must be in the "fastq" subfolder in the working directory:
catcheR_scicatch(
group=c("docker","sudo"),
folder,
expression.matrix,
reference = "GGCGGCTTCATCTGGGGGACGCG",
UCI.length = 6,
threads = 2,
percentage = 15,
ratio = 5,
mode = "bimodal",
x = 100, y = 400)
Example usage:
catcheR_scicatch(group = "docker",
folder = "path/to/working/folder",
expression.matrix = "filename.csv",
threads = 12)
catcheR_scicatch key outputs are the same as for catcheR_10Xcatch (step 2 of iPS2 10X-seq perturbation deconvolution and Figure SP4.4), except that in "silencing_matrix.csv" "cell ID" indicates the PCR well and RT barcode ID (which can be leveraged to identify nuclei from different samples pooled on the same RT plate) Example of cell name after the analysis:
P24__RT_27_7_GCCTGTGT_SCR_ACGGTG
In addition, catcheR_scicatch provides two quality control outputs to evaluate experimental biases during sci-RNA-seq library preparation: demux and RT detail about how many cells were identified from each row and column of the PCR and RT plates, respectively
OPTIONAL: fine tune iPS2-sci-seq perturbation assignment
Run catcheR_scicatch to execute the complete analysis with automatic thresholding; the arguments are the same as for catcheR_10Xcatch (step 32.2), except that no filenames are provided since fastq files must be in the "fastq" subfolder in the working directory:
catcheR_scicatch(
group=c("docker","sudo"),
folder,
expression.matrix,
reference = "GGCGCTTCATCTGGGGGACGCG",
UCI.length = 6,
threads = 2,
percentage = 15,
ratio = 5,
mode = "bimodal",
x = 100, y = 400)
Example usage:
catcheR_scicatch(group = "docker",
folder = "path/to/working/folder",
expression.matrix = "filename.csv",
threads = 12)
catcheR_scicatch key outputs are the same as for catcheR_10Xcatch (step 2 of iPS2-10X-seq perturbation deconvolution and Figure SP4.4), except that in "silencing_matrix.csv", "cell ID" indicates the PCR well and RT barcode ID (which can be leveraged to identify nuclei from different samples pooled on the same RT plate).
Example of cell name after the analysis:
P24__RT_27_7_GCCTGTGT_SCR_ACCGTC
In addition, catcheR_scicatch provides two quality control outputs to evaluate experimental biases during sci-RNA-seq library preparation: demux and RT detail about how many cells were identified from each row and column of the PCR and RT plates, respectively.
Run catcheR_scicatchQC following the steps described for catcheR_10XcatchQC in OPTIONAL: fine tune iPS2-10X-seq perturbation assignment: the functions share the same arguments and outputs, but catcheR_scicatchQC used the output of catcheR_scicatch.
Example usage:
catcheR_scicatchQC(
group = "docker",
folder = "path/to/working/folder",
mode = "noise")
Run catcheR_scinocatch following the steps described for catcheR_10Xnocatch in OPTIONAL: identify cells expressing no shRNA in iPS2-10X-seq: the functions share the same arguments and outputs, but catcheR_scinocatch uses the output of catcheR_scicatch.
Example usage:
catcheR_scinocatch(
group = "docker",
folder = "path/to/working/folder",
expression.matrix = "filename.csv",
threshold = 5,
reference = "TACGCGTTCATCTGGGGGAG")
Barcode reassignment
catcheR_sortcatch is an optional function that corrects the annotated cell-by-gene count matrix obtained with catcheR_10Xcatch or catcheR_scicatch reassigning the perturbation of any cell belonging to a HPSC clone with reliable evidence of a shRNA-barcode swap, based on the results of catcheR_step1QC.
In a new working folder:
- (a) Copy the annotated count matrix (i.e., ”silencing_matrix.csv”)
- (b) Copy the CSV file with the list of UCI-BCs with reliable evidence of a shRNA-barcode swap (i.e., "reliable_clones_swaps.csv”; output 2a of Pooled cloning step 1 QC)
Run catcheR_sortcatch:
catcheR_sortcatch(
group=c("sudo","docker"),
folder,
expression.matrix,
swaps)
catcheR_sortcatch arguments:
- (a) group: string with two options: sudo or docker, depending on the user group
- (b) folder: string with the working folder path
- (c) expression.matrix: string with the filename of the annotated count matrix CSV
- (d) swaps: a character string with the filename of the txt file (step 1b)
catcheR_sortcatch output is an updated annotated gene expression matrix CSV file called "silencing_matrix_updated.csv", in which "BC" and "GENE" have been modified to reflect the actual shRNA encoded in each cell.
Perturbation effect analysis
The second part of catcheR provides an exploratory analysis with visual and statistical summaries to highlight perturbation effects.
Annotation
Before proceeding, the gene expression matrix should be annotated with gene symbols using the scannobyGtf function from the R package rCASC. As part of quality control, we recommend evaluating the fraction of ribosomal and mitochondrial reads — e.g., using the mitoRiboUmi function from the same package — and considering the exclusion of cells with abnormally high proportions, which may indicate poor quality or stress.
Note: After this step, row names of the matrix (the genes) will have the following format:
GeneSymbol:EnsemblID
Example:
ENSG00000000003:TSPAN6
Data loading
In a new working folder:
- Copy the count matrix annotated with gene names during the previous step (filtered_annotated_silencing_matrix_complete_all_samples.csv).
- Copy the file rc_barcodes_genes.csv described in step 1a of Pooled cloning step 2 and hiPSC genome editing QC.
- Create a new-line separated file listing the control genes (e.g. SCR, B2M).
- Create a new-line separated file listing the control samples, if any (e.g. 1,3). These are the sample names also used by aggr (see CSV file used as input for aggr).
- Create a plain text file listing the sample replicate labels, one per line (e.g., batch1, batch1, batch2, batch2). The order of entries must match the order of samples in the input matrix exactly. This file is required for downstream batch-aware analyses.
- If the dataset includes samples from different experiments or processing batches, batch correction is recommended.
- Create a CSV file listing each sample along with its corresponding annotation name, which will be used for display instead of the sample number (this file is mandatory). Example of CSV file to download on GitHub.
- Create a newline-separated file listing the genes of interest whose expression you wish to visualize on the UMAP (optional).
Run catcheR_load:
catcheR_load(
group="docker",
folder,
expression.matrix,
control_genes,
control_samples = NULL,
replicates = NULL,
sample_names,
resolution = 8e-4,
genes = NULL)
Example usage:
catcheR_load(
group="docker",
folder="/path/to/working/folder/",
expression.matrix = "annotated_silencing_matrix_complete_all_samples.csv",
control_genes = "controls.txt",
control_samples = "noTET.txt",
replicates = "replicates.txt",
sample_names = "samples.csv",
resolution = 8e-4,
genes = "genelist.txt")
The argument _resolution_ refers to the resolution parameter used by Monocle’s _cluster_cells_ function for clustering.
catcheR_load outputs:
- 1. expression_data.csv and cell_metadata.csv, which can be used to create a Monocle Cell Data Set (CDS) and are also included in the ready-to-load R object starting_cds.Rdata
- 2. UMAP.pdf plots the dimensionality reduction and UMAPgeneexpression.pdf shows the gene expression on the UMAP of the genes provided by the argument "genes"
- 3. UMAP_clustering.pdf shows the clustering obtained on the UMAP with the provided resolution
- 4. processed_cds.RData is the Cell Data Set after normalization, dimensionality reduction, clustering and calculation of trajectories
At the end of this step, the data are stored in a format compatible with the monocle3 package. However, you can switch to other single-cell analysis frameworks such as Seurat or Scanpy at any point in the workflow. Note: For standard iPS2-seq perturbation analysis, always continue using the CDS object generated by catcheR.
library(SeuratWrappers)
library(Seurat)
seurat = as.Seurat(cds, assay = NULL)
scanpy_sce = as.SingleCellExperiment(seurat)
The follow up analysis are: catcheR_pseudotime, catcheR_modules and catcheR_enrichment.
- With the Pseudotime function, catcheR calculates the cumulative frequency of cells sharing the same gene, shRNA, or clone, compared to each control, along the pseudotime trajectory. catcheR can compare the cumulative frequency curves by computing directional Kolmogorov-Smirnov statistics between the target and control
- With the Genes modules function, catcheR identifies gene modules within the dataset and computes the cumulative frequency of cells with the same gene, shRNA, or clone—relative to each control—based on the expression of each gene module. Also in this case, cumulative frequency curves are compared using a directional Kolmogorov-Smirnov test
- With the Enrichment / depletion analysis function, catcheR calculates the enrichment of cells with the same gene, shRNA, or clone within clusters, in comparison to each control. catcheR can compare each target and control quantity and obtain statistics using the Fisher exact test
Pseudotime
Since Monocle pseudotime calculation requires the user to select the starting point interactively, the first step of the pseudotime analysis needs to be done separately within monocle3.
After loading the processed_cds.RData in R, calculate the pseudotime trajectory and plot it with the R script below:
library(monocle3)
cds = order_cells(cds)
pt = as.data.frame(pseudotime(cds))
names(pt) = c("pseudotime")
write.csv(pt, paste0(dir, "/pseudotime.csv"))
plot_cells(cds, color_cells_by = "pseudotime", label_cellgroups=FALSE, label_leaves=FALSE, label_branch_points=FALSE, graph_label_size=1.5)
Run catcheR_pseudotime:
catcheR_pseudotime( group=c("docker","sudo"), folder, cds, pseudotime, all = FALSE)
Example usage:
catcheR_pseudotime(
group="docker",
folder="/path/to/working/folder/",
cds = "processed_cds.RData",
pseudotime = "pseudotime.csv")
Argument "all" is a logical operator indicating whether to perform the Kolmogorov-Smirnov test against the controls together (all = F) or against each control separately (all = T). The default argument is false
Below is a list of the catcheR_pseudotime outputs:
- (a) The "cumulativefrequencypseudotime" plots show the number of cells at each given point of the pseudotime for different groups of cells at different comparison levels - gene, shRNA, and clone
- (b) The "ks_statistics" CSV files that include the Kolmogorov-Smirnov test results comparing cumulative the frequency based on the pseudotime between knockdown and different controls
- (c) The Volcano plots of the results of the Kolmogorov-Smirnov test show the significance and fold change between the pseudotime cumulative curves. Different controls and levels are used
- (d) The "correlated_pseudotime_gene_exp.pdf" showing the expression on the UMAP of the genes most correlated with the pseudotime
Gene Modules
This function uses Monocle to find gene modules which expression varies in different perturbation groups, i.e., perturbed gene or shRNA or clones.
Run catcheR_modules:
catcheR_modules(group=c("docker", "sudo"), folder, cds, resolution=1e-2)
Example usage:
catcheR_modules(group="docker", folder="/30tb/3tb/data/ratto/testing/", cds = "processed_cds.RData")
The resolution parameter refers to the value used in Monocle’s find_gene_modules function, which influences the number of gene modules identified and the number of genes included in each module (higher resolution typically results in more, smaller modules)
Below are listed the outputs of catcheR_modules:
- (a) gene_modules.csv, listing the genes present in each module
- (b) Heatmap plots display the Z-scores of each module, either all modules or the top 10 most variable ones, across perturbation groups
- (c) The "modulescells" folder contains tables with aggregated module expression values for each cell. These can be used as input for catcheRpseudotime in place of the pseudotime CSV file, allowing the analysis of cumulative frequency based on module expression
Run catcheR_pseudotime on the CSV files stored in the "modules_cells" folder to use the data generated by catcheR_modules as described in the previous example in the Pseudotime section
Enrichment / depletion
This function evaluates whether perturbation groups are enriched or depleted in cells (number) or in specific cell subpopulations (clusters).
Run catcheR_enrichment:
catcheR_enrichment(
group=c("docker","sudo"),
folder,
file,
meta,
timepoint = "PSC",
control_gene = "SCR",
min_cells_cluster = 70,
min_cells_shRNA = 40)
Example usage:
catcheR_enrichment(
group = "docker",
folder = "/3tb/data/ratto/aggr/test/",
file = "processedcds.RData",
meta = "cell_metadata.csv",
timepoint = "PSC",
control_gene = "SCR")
The required input is the cds file generated from catcheR_load. Timepoint refers to the baseline time point used as a control for statistical analysis. It is required when experiments include multiple time points and aim to assess enrichment or depletion across these time points. control_gene specifies the control gene used as a reference for statistical comparisons.
Below are the outputs of catcheR_enrichment:
- (a) Plots of cells in each perturbation group
- (b) Volcano plot showing enrichment or depletion of cell numbers in perturbation groups compared to the control, based on fold-change (log2 ratio) and statistical significance. A corresponding bar plot displays the log2 fold-change for each perturbation
- (c) Barplots showing the distribution of cells from different perturbation groups in the clusters
- (d) Volcano plot showing the results of Fisher’s exact test, comparing the distribution of cells from perturbation groups across Monocle-derived clusters. The plot displays – log10 adjusted p-values versus log2 fold-changes relative to the control group
- (e) Table with Fisher’s statistics