Nov 24, 2021

Public workspaceCell Interaction by Multiplet sequencing (CIM-seq) V.2

DOI
https://dx.doi.org/10.17504/protocols.io.b2byqapw
Cell Interaction by Multiplet sequencing (CIM-seq)
  • Nathanael Andrews1,
  • Jason T. Serviss1,
  • Natalie Geyer (Karolinska Institute Stockholm)1,
  • Agneta B. Andersson1,
  • Ewa Dzwonkowska1,
  • Iva Šutevski1,
  • Rosan Heijboer1,
  • Ninib Baryawno (Karolinska Institute Stockholm)1,
  • Marco Gerling1,
  • Martin Enge1
  • 1Karolinska Institute Stockholm
Nathanael Andrews
Icon indicating open access to content
QR code linking to this content
DOI: https://dx.doi.org/10.17504/protocols.io.b2byqapw
External link: https://www.nature.com/articles/s41592-021-01196-2
Protocol Citation: Nathanael Andrews, Jason T. Serviss, Natalie Geyer (Karolinska Institute Stockholm), Agneta B. Andersson, Ewa Dzwonkowska, Iva Šutevski, Rosan Heijboer, Ninib Baryawno (Karolinska Institute Stockholm), Marco Gerling, Martin Enge 2021. Cell Interaction by Multiplet sequencing (CIM-seq). protocols.io https://dx.doi.org/10.17504/protocols.io.b2byqapwVersion created by Nathanael Andrews
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: November 24, 2021
Last Modified: November 24, 2021
Protocol Integer ID: 55384
Keywords: cell interaction, rnaseq, single cell, cell interactions between every cell type, cell interaction by multiplet, revealing rare cell type, dissociated cell, rare cell type, constituent cell type, cell, cell type, direct physical cell, varying transcriptome, sequencing method, sequencing, genome, rna, specific tissue, computational deconvolution, study of tissue
Abstract
Single cell sequencing methods facilitate the study of tissues at high resolution, revealing rare cell types with varying transcriptomes or genomes, but so far have been lacking the capacity to investigate cell-cell interactions. Here, we introduce CIM-seq, an unsupervised and high-throughput method to analyze direct physical cell-cell interactions between every cell type in a given tissue. CIM-seq is based on RNA sequencing of incompletely dissociated cells, followed by computational deconvolution of these into their constituent cell types using machine learning. CIM-seq is broadly applicable to studies that aim to simultaneously investigate the constituent cell types and the global interaction profile in a specific tissue.
Materials
MATERIALS
ReagentLambda Exonuclease - 5,000 unitsNew England BiolabsCatalog #M0262L
ReagentHotStart ReadyMix (KAPA HiFi PCR kit)Kapa BiosystemsCatalog #KK2601
ReagentpsfTn5addgeneCatalog #79107
Reagent10% SDS solutionTeknovaCatalog #S0287
ReagentSMARTScribe Reverse TranscriptaseTakara Bio Inc.Catalog #634888
ReagentMagnesium chloride solution for molecular biology (1.00 M)Merck MilliporeSigma (Sigma-Aldrich)Catalog #M1028
ReagentTriton X-100 Merck MilliporeSigma (Sigma-Aldrich)Catalog #93426
ReagentKAPA HiFi PCR kit with dNTPsFisher ScientificCatalog #NC0142652
ReagentRecombinant RNAse InhibitorTakara Bio Inc.Catalog #2313A
ReagentBetaine 5MMerck MilliporeSigma (Sigma-Aldrich)Catalog #B0300
ReagentERCC RNA Spike-In MixThermo FisherCatalog #4456740
ReagentUSB Dithiothreitol (DTT), 0.1M SolutionThermo FisherCatalog #707265ML
ReagentUltraPure™ DEPC-Treated WaterThermo FisherCatalog #750023
ReagentdNTP Mix (10 mM each)Thermo FisherCatalog #R0192
ReagentSera-Mag Speed BeadsGE HealthcareCatalog #65152105050250
ReagentHard-Shell® 384-Well PCR Plates thin wall skirtedBio-Rad LaboratoriesCatalog #HSP3801
ReagentQubit 1X dsDNA HS Assay KitThermo Fisher ScientificCatalog #Q33231
ReagentPEG 8000 - (Polyethylene glycol)Merck MilliporeSigma (Sigma-Aldrich)Catalog #P2139
ReagentTAPSMerck MilliporeSigma (Sigma-Aldrich)Catalog #T5130
Oligonucleotides (all ordered from IDT using Standard desalting, except barcodes ordered in solution/plates)

Oligo-dT: AAGCAGTGGTATCAACGCAGAGTACTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT(N1:34333300)(N2:25252525)
IS_PCR: 5′-AAGCAGTGGTATCAACGCAGAGT-3′
TSO: 5′-AAGCAGTGGTATCAACGCAGAGTACATrGrG+G-3′
ME-A: 5'-TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG-3'
ME-B: 5'-GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG-3'
ME-Rev: 5'-/5Phos/CTGTCTCTTATACACATCT-3'

Illumina-compatible barcodes used (Sxxx/Nxxx series, n=784) are available as a supplementary table in the manuscript.
Before start
Before preparing cell lysis plates it is recommended to thoroughly clean all equipment with 70% EtOH and RNAse away to prevent contamination and avoid RNA degradation.
Prepare Lysis buffer
CIM-seq is compatible with both plate based and droplet based single cell RNA-seq methods. The following protocol is for Smart-seq2 based CIM-seq. For droplet based (10x), see the methods section of the manuscript and the vignettes in the Enge lab Github repository.
Prepare Lysis Buffer:
NOTE: Reagents are prepared on ice, working quickly. ERCC is stored in single-use aliquots at Temperature-80 °C , thawed on ice and added last.

Reagent:Reagent concentration:μl per reaction:
H20-1.31
Inhibitor1 U/μl0.05
ERCC (1:600 000)-0.05
Triton-X100 (10% solution)0.2%0.04
10mM dNTP2.5mM/each0.5
100uM dT2.5μM0.05
Total-2
Add Amount2 µL lysis buffer mix to each well. Cover with appropriate lids. Spin down.
Snap freeze on dry ice. Store until use at Temperature-80 °C

Sort cells
Sort single cells and multiplets (aggregates of multiple cells) into Amount2 µL lysis buffer mix.
Multiplets can be discerned from singlets by gating on the basis of FSC-W (Forward scatter - Width) and FSC-H (Forward scatter - Height) (see Figure 1).

Figure 1. Gating scheme and result of FSC-W and FSC-H based sort of multiplets (top) and singlets (bottom) using HCT116 cells .
Following sort, immediately seal with appropriate seals (approved for -80C > 100C) and centrifuge at Centrifigation2000 x g, 4°C, 00:05:00 .
Snap freeze on dry ice. Store until use at Temperature-80 °C .


Reverse transcription and cDNA amplification
Primer annealing
Thaw plate. Spin down. Incubate in thermocycler at Temperature72 °C for Duration00:03:00 . Place on ice immediately.

Prepare RT master-mix

Made fresh.

Reagent:Reaction concentration:Reagent volume:
SmartScribe15u/µl0.475
RNase Inhibitor1.66u/µl0.125
5x First Strand buffer1X1
DTT (100mM)8.33mM0.25
Betaine (5M) [fridge]1.66M1
MgCl2 (1M) [bench]10mM0.03
TSO (100uM)1.66µM0.05
H20-0.07
Total-3

Dispense Amount3 µL per well.
Cover plate with new film and spin down.
Incubate in thermocycler
Temperature42 °C Duration01:30:00
Temperature70 °C Duration00:05:00
Temperature4 °C hold

cDNA preamplification

Made fresh.

ReagentsReaction concentration:Reagent volume:
H2O-1.0688
Kapa HiFi HotStart ReadyMix (2x)1X6.25
IS_PCR primer (10uM)0.1µM0.125
Lambda Exonuclease0,045u/µl0.05625
Total-7.5000
Dispense Amount7.5 µL per well . Total reaction volume will be 12.5µl.
Spin down. Cover with new lid.



Incubate in thermocycler with the following program:
StepTemperatureTimeCycles
Lambda exonuclease37ºC30 min1x
Initial denaturation95ºC3 min1x
Denaturation98ºC20 sec18-24x
Annealing67ºC15 sec
Elongation72ºC4 min
Final elongation72ºC5 min1x
4CHold



cDNA cleanup

We prepare SPRI-beads in 20% PEG-8000 solution as in: https://openwetware.org/wiki/SPRI_bead_mix#Ingredients_for_50_mL_2

Using 20% SPRI-bead solution:

1. Add 0.7x the reaction volume of SPRI beads per well. Mix well by pipetting. (i.e Amount8.75 µL SPRI-bead solution for Amount12.5 µL reaction volume)
2. Incubate Duration00:05:00 TemperatureRoom temperature
3. Place on magnetic stand for Duration00:03:00
4. Carefully remove supernatant
5. Add 40 µl 80% EtOH and incubate Duration00:00:30
6. Remove EtOH (without disturbing the beads)
7. Wash again with EtOH. Make sure to remove well.
8. Allow beads to air-dry for Duration00:10:00 -Duration00:15:00
9. Remove plate from magnetic stand
10. Elute beads in Amount15 µL EB or TE buffer. Mix well by pipetting
11. Incubate Duration00:05:00 TemperatureRoom temperature
12. Place on magnetic plate for Duration00:03:00
13. Optional: Carefully remove supernatant to the elution plate
cDNA quantification

We measure concentration of random wells using Qubit HS dsDNA, adapted to a 96-well plate reader.

1. Add Amount97 µL of 1X Qubit HS dsDNA solution to a flat-bottom, black plate
2. Add Amount3 µL of cDNA sample
3. Add Standards (NOTE: We make a 8-step ladder from 0ng/µl --> 10ng/µl Qubit Standard DNA in TE buffer)
3. Read in plate reader using 485nM excitation/528nm emission
4. Calculate cDNA concentration
(optional) cDNA quality control

Using Agilent HS 5000 DNA chips (or equivalent)


Figure 2. cDNA profile of single cell run on HS D5000 Agilent tapestation

Make cDNA dilution plate

Dilute cDNA in water based on average concentration from Qubit measurements.
Target concentration 150pg per µl in Amount15 µL .

cDNA tagmentation
Tn5 digestion


Tn5 is produced from psfTn5 (Addgene #79107), purified to ~3mg/ml and assembled with Illumina Tn5 adapters (see oligos) as in Picelli et al, 2014.
Citation
Picelli S, Björklund AK, Reinius B, Sagasser S, Winberg G, Sandberg R (2014). Tn5 transposase and tagmentation procedures for massively scaled sequencing projects. Genome research.
https://doi.org/10.1101/gr.177881.114
LINK

Prepare Tn5 master mix

Note
NOTE: TAPS-PEG Buffer contains PEG, which is viscous. Buffer should equilibrate to room temperature before use to allow proper mixing.


ReagentReaction conc.µl per reaction
Nuclease free H2O-1.05
TAPS-PEG (50mM TAPS, 25mM MgCl2, 40% PEG-8000)10mM TAPS, 5mM MgCl2, 8% PEG-80000.5
psfTn5, loaded with 50µM MEDS-A/B0.25
Total1.8
Dispense Amount1.8 µL per well in a new plate(tagmentation plate)

Add Amount0.7 µL cDNA (normalized to 150pg/µl)
Mix well by vortexing plate. Cover with new lid and spin down.
Incubate in thermocycler at Temperature55 °C Duration00:10:00
Remove immediately and stop reaction by adding Amount1 µL per well of 0.1% SDS.
Vortex, spin down and incubate Duration00:07:00 at TemperatureRoom temperature

cDNA library PCR and barcoding
Make PCR master-mix

Reagentsµl per reaction
H2O13.25
5x buffer5
dNTPs0.75
KAPA0.5
Total19.50
Dispense Amount19.5 µL per well to tagmentation plate (containing Amount3.5 µL sample after step 14)
Add primers/barcodes

Amount2 µL per well (from 384-well index plates, with 3.75µM/each forward/reverse primers; see oligos in materials).

Total reaction volume is Amount25 µL (Amount3.5 µL sample + Amount19.5 µL PCR mix and Amount2 µL primers).

Vortex. Spin down and cover. Incubate in thermocycler as following:
StepTemperatureTimeCycles
Gap fill72ºC3 min1x
First denature95ºC30 sec1x
Denature95ºC15 sec12x
Denature67ºC30 sec
Denature72ºC45 sec
Final extension72ºC4 min1x
4-10ºChold

Pool Amount2.5 µL from each well to an 1.5ml Eppendorf tube.
Library cleanup

1. Add 0.9x pooled library volume of SPRI-bead solution. Incubate for Duration00:05:00 at TemperatureRoom temperature .
2. Place on magnetic rack for Duration00:03:00 .
3. Remove supernatant without disturbing magnetic beads.
4. Add at least Amount1 mL 80% EtOH (fresh). Incubate for Duration00:00:30 ..
5. Remove supernatant.
6. Repeat EtOH wash.
7. Air dry for Duration00:10:00 - Duration00:15:00 .
8. Re-suspend beads thoroughly in 100 µl EB or TE buffer.
9. Place eppendorf on magnetic rack for Duration00:03:00 .
10. Transfer supernatant to new 1.5ml Eppendorf tube.
11. Repeat cleanup (from step 1-7) and elute in 30 µl EB or TE buffer.
12.(Optional) Place eppendorf on magnetic rack for Duration00:03:00 and transfer supernatant to new tube.
Pooled library QC

Figure 3. cDNA profile of a library of 784 cells (both single cells and multiplets) on HS D5000 Agilent tapestation.
Data pre-processing

A series of pre-processing steps must be performed in order to generate a counts file:

1. Trim reads, remove adapter sequences and align RNAseq data to reference genome using STAR:
Citation
Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, Batut P, Chaisson M, Gingeras TR (2012). STAR: ultrafast universal RNA-seq aligner. Bioinformatics (Oxford, England).
https://doi.org/10.1093/bioinformatics/bts635
LINK

2. Remove duplicate reads using Picard:
Citation
McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, Garimella K, Altshuler D, Gabriel S, Daly M, DePristo MA (2010). The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome research.
https://doi.org/10.1101/gr.107524.110
LINK
3. Generate transcript counts file using HTSeq:
Citation
Anders S, Pyl PT, Huber W (2015). HTSeq--a Python framework to work with high-throughput sequencing data. Bioinformatics (Oxford, England).
https://doi.org/10.1093/bioinformatics/btu638
LINK

Dimensionality reduction and classification


Once a counts file has been generated the data can be analyzed. CIM-seq requires four arguments in order to run:

1. The raw counts data with gene IDs as rownames and sample IDs as
colnames.

2. The ERCC spike-in counts data with gene IDs as rownames and sample
IDs as colnames.

3. The dimensionality reduced representation of the data.

4. A class for each of the individual singlets.

In order to generate the last two of these we recommend using the Seurat package in R, as CIM-seq is implemented in R as well. A number of tutorials for Seurat can be found on the Satijalab website:

https://satijalab.org/seurat/vignettes.html

CIM-seq


CIM-seq can be downloaded from:
https://github.com/EngeLab/CIMseq

Or installed directly in R using the devtools package:

devtools::install_github("EngeLab/CIMseq")

The CIM-seq vignette can be found at:

https://github.com/EngeLab/CIMseq/tree/master/vignettes

Citations
Step 13
Picelli S, Björklund AK, Reinius B, Sagasser S, Winberg G, Sandberg R. Tn5 transposase and tagmentation procedures for massively scaled sequencing projects.
https://doi.org/10.1101/gr.177881.114
Step 23
Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, Batut P, Chaisson M, Gingeras TR. STAR: ultrafast universal RNA-seq aligner.
https://doi.org/10.1093/bioinformatics/bts635
Step 23
McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, Garimella K, Altshuler D, Gabriel S, Daly M, DePristo MA. The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data.
https://doi.org/10.1101/gr.107524.110
Step 23
Anders S, Pyl PT, Huber W. HTSeq--a Python framework to work with high-throughput sequencing data.
https://doi.org/10.1093/bioinformatics/btu638