DNA methylation protocol for analyzing cell-free DNA in the spent culture medium of human preimplantation embryos

Summary

Cell-free DNA (cfDNA) in spent embryo culture media (SECM) provides prospects for noninvasive preimplantation genetic testing. Here, we present a post-bisulfite-adapter-tagging (PBAT)-based whole-genome DNA methylation sequencing protocol (SECM-PBAT) for human SECM cfDNA analysis. We describe steps for SECM lysis, bisulfite conversion and purification, preamplification by random priming, tagging adapter II, and library establishment. We then detail library quality control, sequencing, and bioinformatics analysis. This approach simultaneously detects chromosome aneuploidy and deduces the proportional contributions of cellular components.

For complete details on the use and execution of this protocol, please refer to Chen et al. (2021).¹

Graphical abstract

Highlights

• •Noninvasive preimplantation genetic testing
• •Whole-genome DNA methylation sequencing for cfDNA in spent embryo culture media
• •Simultaneously detecting chromosome aneuploidy and cellular origins of cell-free DNA
• •SECM cfDNA comprises the contributions of blastocysts, cumulus cells, and polar bodies

Publisher’s note: Undertaking any experimental protocol requires adherence to local institutional guidelines for laboratory safety and ethics.

Cell-free DNA (cfDNA) in spent embryo culture media (SECM) provides prospects for noninvasive preimplantation genetic testing. Here, we present a post-bisulfite-adapter-tagging (PBAT)-based whole-genome DNA methylation sequencing protocol (SECM-PBAT) for human SECM cfDNA analysis. We describe steps for SECM lysis, bisulfite conversion and purification, preamplification by random priming, tagging adapter II, and library establishment. We then detail library quality control, sequencing, and bioinformatics analysis. This approach simultaneously detects chromosome aneuploidy and deduces the proportional contributions of cellular components.

Before you begin

In recent years, the cell-free DNA (cfDNA) in spent embryo culture media (SECM) has attracted attention as it represents a good source of sampling for noninvasive preimplantation genetic testing (niPGT) that avoids the potential safety concerns of the widely applied trophectoderm (TE) biopsy.^{2,3,4,5,6,7,8} We have recently established a post-bisulfite adapter tagging (PBAT)-based whole-genome DNA methylation sequencing method (SECM-PBAT) for SECM cfDNA analysis.¹ It is adapted from the single-cell PBAT method that we have applied for DNA methylation analysis of human preimplantation embryos.⁹ Using the method, we have demonstrated that not only blastocysts and cumulus cells, but also polar bodies, contribute to the SECM cfDNA.¹ The contribution of polar bodies to SECM cfDNA has been previously proposed by others.^5,6,8,10 The method is also able to detect chromosome aneuploidy. By integrating the chromosome aneuploidy and cellular proportion information, the SECM-PBAT method shows prospects for noninvasive preimplantation genetic testing for aneuploidy (niPGT-A).

Institutional permissions

All experiments involving human samples must be approved by institutional permissions and national laws and regulations. This study was approved by the Reproductive Medicine Ethics Committee of Peking University Third Hospital (Research License 2019-393-02).

Embryo culture and SECM collection

Timing: 6–8 days

• 1.
  Before starting the protocol, culture the embryo in vitro and collect the spent culture medium.

  Note: The whole process should be operated in a sterile in vitro fertilization (IVF) workstation (e.g., IVF Workstation, CooperSurgical).

  • a.
    Cover the embryo culture medium with mineral oil and then prewarm and equilibrate at 37°C with 6% CO₂ and 5% O₂ in the incubator 10 h–12 h before use. Prepare embryo transfer pipettes (inner diameter of ∼200–250 μm) by pulling glass Pasteur pipettes and then sterilize them.
  • b.
    Denude cumulus cells of the oocyte-corona-cumulus complex (OCCCs) with hyaluronidase (80 IU/mL) and stripper pipettes.

    Note: It is acceptable that a few cumulus cells still attach to the oocyte, which can be more completely removed on day 3.

    • i.
      Culture the oocyte in the G-IVF Plus medium.
    • ii.
      Perform intracytoplasmic sperm injection (ICSI), a micromanipulation in which a single sperm is injected into a metaphase-II oocyte, to avoid sperm contamination (Figures 1A–1C).

      Note: In a IVF cycle, the day for oocyte retrieval and fertilization is defined as day 0, and use the G-MOPS Plus medium during the ICSI procedure.

    • iii.
      Culture the oocyte in the G-1 Plus medium.

      Note: Further details on ICSI have been previously described.¹¹

  • c.
    On day 1, assess whether fertilization is successful at approximately 16–18 h after ICSI (Figure 1D). Culture the zygotes until day 3 in the G-1 Plus medium.
  • d.
    On day 3, check the embryos under a microscope to examine whether there remain some cumulus cells or not (Figures 1E and 1F), culture the embryos with the G-2 Plus culture medium.

    Note: For embryos with residual cumulus cells, gently remove the cumulus cells using an appropriate stripper. Transfer the embryos to a new blastocyst culture dish with the G-2 Plus culture medium.

    Note: Different embryos from one couple can be group cultured in the same drop of 20–50 μL culture medium per drop from day 0 to day 3.

  • e.
    On day 4, transfer each embryo to a separate new drop of 15–20 μL G-2 Plus culture medium in a new blastocyst culture dish (Figure 1G).
  • f.
    From day 5–7, check each embryo under a microscope every day.

    • i.
      When an embryo reaches the fully expanded blastocyst stage (Figure 1H), remove the embryo to another dish.
    • ii.
      Collect the corresponding SECM in an individual polymerase chain reaction (PCR) tube (Figure 1I).

      Note: The samples can be stored in a refrigerator at −20°C for up to one month.

Figure 1.

ICSI, embryo culture and SECM collection

(A) An oocyte-corona-cumulus complex at day 0 of IVF. Scale bar: 40 μm.

(B) An oocyte after removal of cumulus cells. Scale bar: 40 μm.

(C) An oocyte undergoing ICSI operation. Scale bar: 40 μm.

(D) A fertilized zygote at day 1 of IVF. Scale bar: 40 μm.

(E) A cleavage embryo with a few cumulus cells at day 3 of IVF. Scale bar: 40 μm.

(F) A cleavage embryo with the residual cumulus cells being removed at day 3 of IVF. Scale bar: 40 μm.

(G) A blastocyst culture dish with each single embryo being cultured individually from day 4 on.

(H) A expanded blastocyst. Scale bar: 40 μm.

(I) The SECM samples are collected in PCR tubes.

Preparing buffers

Timing: 1–4 h

Before starting the protocol, make sure that the reagents are prepared and available in sufficient volume.

• 2.
  Prepare CT conversion reagents according to the DNA Methylation-Direct MagPrep kit instruction.

  • a.
    Carefully open the CT conversion reagent bottle to prevent the powder from floating out.
  • b.
    Add 7.9 mL M-Solubilization Buffer and 3 mL M-Dilution Buffer to the bottle, and vortex thoroughly at 20°C–25°C for 15 min to ensure that the powder of CT conversion reagent is fully dissolved.
  • c.
    Add 1.6 mL M-Reaction Buffer and thoroughly mix by vortexing for 5 min.

Note: The CT conversion reagents should be stored carefully away from light, and can be stored at −20°C for one month. Please reduce repeated freezing and thawing, which will damage the CT conversion reagents.

• 3.
  Prepare Proteinase K according to the DNA Methylation-Direct MagPrep kit instruction.

  • a.
    Add 1,040 μL of Proteinase K Storage Buffer to the tube containing Proteinase K powder. Then store at −20°C.
• 4.Prepare the M-Wash Buffer and the M-Elution Buffer according to the DNA Methylation-Direct MagPrep kit instruction.

Note: The M-Wash Buffer can be stored at 20°C–25°C for one year. Before use, add 4 volumes of 100% ethanol to the M-Wash Buffer.

Preparing primers and barcodes

Timing: 3 days

• 5.
  Design and synthesize adapters and primers in advance.

  • a.
    Centrifuge the oligonucleotide powder at full speed (>12,000g) for 2 min, add nuclease-free water to 100 μM for the master stock, and incubate at 37°C for 30 min.
  • b.
    Dilute primers with nuclease-free water for the random priming and tagging processes to 10 μM, and for the PCR amplification process to 15 μM as working stocks.
  • c.
    Store all master and working stocks at −20°C until further use.

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Chemicals, peptides, and recombinant proteins
G-MOPS PLUS	Vitrolife	10130
G-IVF PLUS	Vitrolife	10136
G-1 PLUS	Vitrolife	10128
G-2 PLUS	Vitrolife	10132
Mineral oil	Vitrolife	10029
Hyaluronidase	Sigma-Aldrich	H4272
Tris-EDTA	Sigma-Aldrich	T9285
Triton X-100	Sigma-Aldrich	T8787
1 M KCl	Sigma-Aldrich	60142
Lambda DNA	New England Biolabs	N3011S
Proteinase K	Zymo	D3001-2-20
dNTP Mixture	TaKaRa	4019
Klenow DNA polymerase (3′–5′ exo-, 50 U/μL)	TIANGEN	NG202-02
Klenow DNA polymerase (3′–5′ exo-, 50 U/μL)	New England Biolabs	M0212L
Blue Buffer (10×)	TIANGEN	NG202-01
AMPure XP Beads	Beckman Coulter	A63882
Nuclease-free water	Ambion	AM9932
Ethanol	Sigma-Aldrich (or other molecular biology grade ethanol)	1085430250
Critical commercial assays
DNA Methylation-Direct MagPrep	Zymo	D5044
KAPA HiFi HotStat ReadyMix	Kapa Biosystem	KK2602
Fragment Analyzer DNA/NGS Kits	Agilent	DNF-474-0500
Qubit™ dsDNA HS Assay Kit	Invitrogen	Q32854
Biological samples
SECM samples of human preimplantation embryos	Peking University Third Hospital	N/A
Deposited data
HRA000332	This paper; raw data	https://bigd.big.ac.cn/gsa-human/
PBAT_test_S1.bed.gz	This paper; analyzed data	https://github.com/jasminexiao/niPGT
C_DMRs.bed	This paper; cumulus-specific regions	https://github.com/jasminexiao/niPGT
O_DMRs.bed	This paper; oocyte/polar-body-specific regions	https://github.com/jasminexiao/niPGT
XX_NG_ICM_E_4M.median.bed	This paper; reference sample for Gingko	https://github.com/jasminexiao/niPGT
Oligonucleotides
P5-N9-oligo1	Sangon Biotech	CTACACGACGCTCTTCC GATCTNNNNNNNNN
P7-N9-oligo2	Sangon Biotech	AGACGTGTGCTCTTCCGAT CTNNNNNNNNN
Universal PCR Primer	Sangon Biotech	AATGATACGGCGACCACCG AGATCTACACTCTTTCCCT ACACGACGCTCTTCCGATC
NEBNext® Multiplex Oligos for Illumina® (Index Primers Set 1)	New England Biolabs	E7335L
Software and algorithms
Samtools	Li et al.12	http://samtools.sourceforge.net/
bowtie2	Langmead and Salzberg13	http://bowtie-bio.sourceforge.net/bowtie2/index.shtml
scBS-map	Wu et al.14	https://github.com/wupengomics/scBS-map
bedtools	Quinlan and Hall15	https://bedtools.readthedocs.io/
Gingko	Garvin et al.16	https://github.com/robertaboukhalil/ginkgo
Particle Swarm Optimization	Parsopoulos and Vrahatis17	https://cran.r-project.org/web/packages/pso
Others
DNA LoBind Tubes, 1.5 mL	Eppendorf	022431021
DNA LoBind Tubes, 2.0 mL	Eppendorf	022431048
0.2 mL PCR strip tubes	Axygen	PCR-0208-CP-C
Magnetic stand	Diagenode	kch-816-001
3.5 cm culture plates	Thermo Scientific	150255
Stripper pipettes	CooperSurgical	7-72-4145
Micro injection pipettes	Cook Medical	G26684
Holding micro pipettes	CooperSurgical	MPH-MED-35
IVF workstation	CooperSurgical	L126
Fragment analyzer	Advanced Analytical Technologies	FSv2-CE
Thermomixer	Eppendorf	5355 000.011
PCR thermocycler	Analytik Jena	Biometra TRIO
Sequencer	Illumina	NovaSeq 6000
Low-binding filter tips 10 μL	QSP	TF104-10-Q
Low-binding filter tips 20 μL	QSP	TF113-20-Q
Low-binding filter tips 200 μL	Bioscience	BLD-BS1007200
Low-binding filter tips 1000 μL	Kirgen	KG5311

Materials and equipment

10% Triton X-100

Reagent	Amount	Final concentration
Triton-X 100	1 mL	10%
nuclease-free water	9 mL	N/A
Total	10 mL	N/A

Store at 20°C–25°C for up to 1 year.

Lysis buffer

Reagent	Amount (per sample)	Final concentration
1 M Tris-EDTA	0.4 μL	20 mM
1 M KCl	0.4 μL	20 mM
10% Triton-X 100	0.6 μL	0.3%
20 mg/mL protease K	1 μL	1 mg/mL
Lambda DNA (0.06 pg/μL)	1 μL	0.06 pg
Total	3.4 μL

Store at −80°C for up to 1 year.

Step-by-step method details

SECM lysis

Timing: 2–3 h

In this step, cfDNAs of SECM are released after degradation of proteins and other components.

• 1.Measure the volume of each SECM sample with a micropipette and add nuclease-free water to reach a total volume of 16.6 μL. If the original volume of the sample is more than 16.6 μL, use only 16.6 μL.

Note: As the volumes of the collected SECM samples are generally more than 5 μL with variation, we have adapted the lysis step of the single-cell PBAT protocol for SECM by increasing the total volume of the lysis reaction from 5 μL to 20 μL.

CRITICAL: The volume of the SECM sample should be accurately measured.

• 2.
  Add the lysis buffer (see materials and equipment) to each SECM sample from step 1 to make a final lysis reaction volume of 20 μL.

  • a.
    Mix well by vortexing and spinning down for a few seconds.
  • b.
    Immediately place the tube in a heating block, incubating for 1 h at 50°C and then 30 min at 75°C.
• 3.Centrifuge at 9,000 g for 1 min and immediately place the tube on ice.

Pause point: Lysed SECM samples can be stored at −80°C for up to six months.

Bisulfite conversion and purification

Timing: 5–7 h

In this step, unmethylated Cs are changed to Us while methylated Cs (most in the CpG context) are preserved, and thus the information on DNA methylation is provided at single base resolution.

• 4.Dissolve the CT Conversion Reagent by shaking it in a Thermomixer at 37°C until completely clear.

CRITICAL: The CT Conversion Reagent should be stored away from light and fully dissolved when used. The temperature can be increased to 42°C to ensure no precipitation.

• 5.Add 130 μL prepared CT conversion reagent to each 20 μL SECM lysate. Mix by vortexing. Incubate at 98°C for 8 min, 64°C for 3.5 h for bisulfite conversion of the SECM cfDNA, and then hold at 4°C.

Pause point: The bisulfite-converted SECM cfDNA can be stored at −20°C for up to 3 days or 4°C for 1 day.

Troubleshooting.

• 6.
  Purify the bisulfite-converted SECM cfDNA using the Zymo EZ-96 DNA Methylation-Direct™ MagPrep kit.

  • a.
    Add 10 μL of MagBinding Beads and 600 μL of M-Binding Buffer to 1.5 mL DNA LoBind Tubes.
  • b.
    Add the 150 μL bisulfite-converted SECM cfDNA from step 5 to the MagBinding Beads and M-Binding Buffer, vortexing fully for 30 s.
  • c.
    Incubate the mixture at 20°C–25°C for 5 min to bind the DNA to the beads.
  • d.
    Place the tubes on a magnetic stand until the solution clears, and then remove and discard the supernatant by pipetting.
  • e.
    Remove the tubes from the magnetic stand and add 400 μL M-Wash Buffer. Mix the beads by vortexing.
  • f.
    Place the tube on the magnetic stand until the solution clears, and then remove the supernatant by pipetting and discard the supernatant.
  • g.
    Remove the tubes from the magnetic stand and add 200 μL of M-Desulfonation Buffer, and mix thoroughly by vortexing for 30 s.
  • h.
    Let tubes stand at 20°C–25°C for 15 min. Then place the tubes on the magnetic stand until the solution clears, and then discard the supernatant by pipetting.
  • i.
    Remove the tubes from the magnetic stand and add 400 μL of M-Wash Buffer to the beads.
  • j.
    Place the tubes on the magnetic stand until the solution clears, and then discard the supernatant by pipetting. Repeat this step once.
  • k.
    Transfer the tubes to the heating block and dry the beads for 20 min at 55°C.
  • l.
    Resuspend the beads with 21.5 μL of M-Elution Buffer, and heat the elution at 55°C for 2 min to fully elute the DNA.
  • m.
    Place the tubes on the magnetic stand until the solution clears, and then remove and transfer 19.5 μL supernatant to a clean PCR tube.

Note: As the MagBinding Beads precipitate very easily, keep the solution well mixed throughout the process. One can use the preamplification priming mixture for dissolving the DNA, as described by Clark et al.¹⁸

Pause point: Purified bisulfite-converted DNA can be stored at −20°C for no more than two weeks.

Preamplification by random priming

Timing: 4 h

In this step, four rounds of DNA stands (the first strands) complementary to the bisulfite-converted DNA are synthesized using Klenow DNA polymerase and random primers with the adapter I being tagged.

• 7.Prepare the preamplification priming mixture as follows:

Reagent	Amount	Final concentration
Blue Buffer (10×)	2.5 μL	1×
P5-N9-oligo1 (10 μΜ)	1 μL	0.4 μΜ
dNTP Mixture (10 mM)	1 μL	0.4 μΜ
bisulfite-converted DNA	19.5 μL
Total	24 μL

• 8.Mix well by gently pipetting up and down. Incubate the mixture on a heating block at 65°C for 3 min. And then immediately transfer the PCR tubes on ice.

Note: After adding 1 μL of Klenow DNA polymerase in step 9, the total volume will reach 25 μL.

CRITICAL: The tubes need to be sufficiently cooled to avoid reduction of the activity of the Klenow DNA polymerase.

• 9.Add 1 μL of Klenow DNA polymerase (3′-5′ exo-, 50 U/μL, from TIANGEN or New England Biolabs) to the mixture. Mix well by gently pipetting up and down 15 times. Centrifuge at 300 g for 10 s at 20°C–25°C and then place the tubes on a pre-cooled (4°C) thermocycler.
• 10.Incubate the tubes in the thermocycler for one round of preamplification as follows:

Steps	Temperature	Time	Ramp speed (°C/s)
Annealing	4°C	5 min	-
Annealing	4°C–37°C	8.25 min	1
Extension	37°C	30 min	-
Hold	4°C	forever

• 11.Heat the mixture to 95°C for 1 min in a thermocycler and then immediately cool it at 4°C.

CRITICAL: The tubes need to be sufficiently cooled to avoid reduction of the activity of the Klenow DNA polymerase.

• 12.Add the following freshly prepared solution for the next round of preamplification:

Reagent	Amount	Final concentration
dNTP Mixture (2.5 nM)	1 μL	1 nM
P5-N9-oligo1 (25 pΜ)	1 μL	10 pΜ
Klenow (exo-, 50 U/μL)	0.5 μL
Total	2.5 μL

Note: The dNTPs mixture of 2.5 nM is diluted from the dNTPs mixture of 10 mM, and the P5-N9-oligo1 of 25 pΜ is diluted from the P5-N9-oligo1 of 10 μΜ, both with 1× Blue Buffer.

• 13.Mix well by gently pipetting up and down. Incubate the tubes in the thermocycler as follows:

Steps	Temperature	Time	Ramp speed (°C/s)
Annealing	4°C	5 min	-
Annealing	4°C–37°C	8.25 min	1
Extension	37°C	30 min	-
Hold	4°C	forever

• 14.Repeat steps 11–13 for additional two times, making four rounds of preamplification in total.
• 15.
  After the PCR cycles, centrifuge the tubes at 500 g for 1 min. Next, purify the First-strand random priming product using 0.8× AMPure XP beads.

  • a.
    Add 26 μL of AMPure XP Beads to the PCR Tube.
  • b.
    Gently shake the PCR tube until a homogeneous solution is seen.
  • c.
    Incubate the PCR tubes for 10 min at 20°C–25°C for maximum recovery.
  • d.
    Place PCR tubes on the magnetic stand at 20°C–25°C for 5 min until the solution becomes clear.
  • e.
    Remove and discard the supernatant carefully.
  • f.
    Add 200 mL of 70% ethanol to each tube and pipet up and down 10 times gently, then remove the supernatant.
  • g.
    Repeat step f.
  • h.
    Remove the tube from the magnetic stand and place it at 20°C–25°C until the ethanol is evaporated.

    Note: In this step, XP beads become light brown, which means they are in a dry state.

  • i.
    Add 21.5 μL nuclease-free water, and mix by vortexing for 30 s.
  • j.
    Incubate at 20°C–25°C for 2 min and then place the PCR tube again in the magnetic stand until the solution becomes clear.
  • k.
    Transfer 19.5 μL of the supernatant to the next step.

    Note: XP beads need to be placed at 20°C–25°C for 30 min before use.

Tagging adapter II

Timing: 3 h

In this step, the second DNA strand complementary to the first DNA strand is synthesized using Klenow DNA polymerase and random primers with the adapter II being tagged.

• 16.Prepare the adapter II tagging mixture on ice as follows:

Reagent	Amount	Final concentration
10× Blue Buffer	2.5 μL	1×
P7-N9-oligo2 (10 μΜ)	1 μL	0.4 μΜ
dNTP Mixture (10 mM)	1 μL	0.4 μΜ
Purified DNA from step 15	19.5 μL
Total	24 μL

Note: After adding 1 μL of Klenow DNA polymerase in step 18, the total volume will reach 25 μL.

• 17.Heat the mixture at 95°C for 3 min in a PCR thermocycler, then immediately place the PCR tubes on ice.
• 18.Add 1 μL of Klenow DNA polymerase (3′–5′ exo-, 50 U/μL) to the mixture. Mix well by gently pipetting and then centrifuged at 300 g for 10 s.
• 19.Incubate the tubes in the thermocycler for tagging adapter II as follows:

Steps	Temperature	Time	Ramp speed (°C/s)
Annealing	4°C	5 min	–
Annealing	4°C–37°C	8.25 min	1
Extension	37°C	90 min	–
Hold	4°C	Forever

• 20.Use 0.8× AMPure XP beads to purify the second-strand synthesis product referring to step 15, adding 13.5 μL nuclease-free water for elution and transferring 11.5 μL for the next step.

PCR amplification

Timing: 2 h

In this step, the final library product is PCR amplified using primers against adapter I and II.

• 21.Prepare the PCR amplification mixture on ice as follows:

Reagent	Amount	Final concentration
2× KAPA HiFi HotStart ReadyMix	12.5 μL	1×
Index Primer (15 μΜ)	1 μL	0.6 μΜ
Universal PCR Primer (15 μΜ) (from Illumina)	1 μL	0.6 μΜ
Purified DNA from step 20	11.5 μL
Total	25 μL

• 22.Mix thoroughly by vortexing 30 s, centrifuge the tubes at 300 g for 10 s, and perform PCR amplification using the following program:

Steps	Temperature	Time	Cycles
Initial Denaturation	95°C	3 min	1
Denaturation	98°C	20 s	16 cycles
Annealing	65°C	30 s
Extension	72°C	1 min
Final extension	72°C	1 min	1
Hold	4°C	forever

• 23.Centrifuge the tubes at 300 g for 1 min. Purify the PCR products twice using 0.8× AMPure XP beads referring to step 15, add 27 μL nuclease-free water for elution and transfer 25 μL eluate to a new tube each time.

Pause point: The purified PCR products can be stored at −20°C for one month.

Library quality control

Timing: 2 h

In this step, the library is examined for concentration and size distribution as quality control.

• 24.Quantify the concentration of the PCR products from step 23 using a Qubit dsDNA HS assay kit.

Troubleshooting.

• 25.
  Assess the sizes of the PCR products from step 23 using a fragment analyzer equipment.

  • a.
    Dilute the PCR products to a concentration of 1–2 ng/μL.
  • b.
    Add 2 μL of diluted PCR products to 22 μL of Diluent Marker, and then mix vigorously and centrifuge at 500 g for 1 min.
  • c.
    Run the diluted PCR products on the fragment analyzer.

Note: Run the samples immediately after preparation. If the samples have not been used right away, cover and keep them at 4°C for up to 3 days, and warm them to 20°C–25°C and centrifuge at 300 g for 30 s before running.

Troubleshooting.

Sequencing

Timing: 1–8 days

In this step, the library is processed for high-throughput sequencing.

In this step, pooled libraries with compatible indexes are sequenced on Illumina platforms.

• 26.Carry out 150 bp paired-end sequencing on NovaSeq 6000 (Illumina) or other platforms according to the manufacturer’s protocol.

Data processing

Timing: 6–8 h

In this step, the data are processed for mapping to the reference genome and the information on DNA methylation levels at individual cytosine site is obtained.

• 27.Installing system requirements of the scBS-map toolkit.

Note: ScBS-map toolkit can be downloaded from https://github.com/wupengomics/scBS-map. Click in the green ‘Code’ tab for downloading the zip folder, which contains a workflow figure, perl scripts including qcreads.pl, align-end2end.pl, align-local.pl, qcbam.pl, mergebam.pl, and a ReadMe file. Install system requirements including SAMtools, bowtie2 and BS-Seeker2 according to the ReadMe file.

• 28.
  Launching scBS-map toolkit to process raw data.

  • a.
    Remove sequencing adapters, amplification primers, and low-quality bases from the raw bisulfite sequencing reads using qcreads.pl in the scBS-map toolkit. On the PowerShell Prompt Console, type command as follows:

    > perl qcreads.pl -f Sample.R1.fastq.gz -l 10 -o Sample_R1_val_1.fq.gz

    > perl qcreads.pl -f Sample.R2.fastq.gz -l 10 -o Sample_R2_val_2.fq.gz
  • b.
    Discard R2 reads that have more than 3 unmethylated CHs (H = A, T or C), and the corresponding R1 reads are also discarded. Type command as follows:

    > perl ch3deleate.pl Sample_R2_val_2.fq.gz Sample_ R2_clean.fq.gz

    > zcat Sample_ R2_clean.fq.gz | grep 'ˆ@' | awk '{print $1}' >> R2_head.txt

    > perl extract_R1.pl R2_head.txt Sample_R1_val_1.fq.gz Sample_ R1_clean.fq.gz

    Note: The scripts ch3deleate.pl and extract_R1.pl can be found at https://github.com/jasminexiao/niPGT.

  • c.
    The clean reads are first mapped to the human reference genome (hg19) in the end-to-end alignment mode using align-end2end.pl. Type command as follows:

    > perl align-end2end.pl Sample_ R1_clean.fq.gz -g hg19.genome.fa -p 5 -u Sample _R1.unaligned.fq -o Sample_R1.end2end.bam

    > perl align-end2end.pl Sample_ R2_clean.fq.gz -g hg19.genome.fa -p 5 -u Sample _R2.unaligned.fq -o Sample_R2.end2end.bam
  • d.
    Carry out local alignment to unaligned reads using align-local.pl. Type command as follows:

    > perl align-local.pl -f Sample_R1.unaligned.fq -g hg19.genome.fa -p 5 -o Sample_R1.local.bam

    > perl align-local.pl -f Sample_R2.unaligned.fq -g hg19.genome.fa -p 5 -o Sample_R2.local.bam
  • e.
    Remove low-confidence local-aligned reads using qcbam.pl. Type command as follows:

    > perl qcbam.pl -f Sample_R1.local.bam -n 10 -p 4 -o Sample_R1.local.hc.bam

    > perl qcbam.pl -f Sample_R2.local.bam -n 10 -p 4 -o Sample_R2.local.hc.bam
  • f.
    Merge all alignments from end-to-end and local mapping mode using mergebam.pl. Type command as follows:

    > perl mergebam.pl -e Sample_R2.end2end.bam -l Sample_R2.local.hc.bam -p 4 -o Sample_R2.merge.bam

    > samtools merge -f Sample.bam Sample_R1.merge.bam Sample_R2.merge.bam

    > samtools sort Sample.bam Sample.sort

    > samtools index Sample.sort.bam
  • g.
    Remove PCR duplicates using the Picard tools (https://broadinstitute.github.io/picard/). Type command as follows:

    > java -Xmx4g -jar picard.jar MarkDuplicates I=Sample.sort.bam O=Sample.rmdup.bam M=Sample.rmdup.txt CREATE_INDEX=true ASSUME_SORTED=true VALIDATION_STRINGENCY=SILENT REMOVE_DUPLICATES=true 2> $output/$sample/${sample}.picard.log

    samtools index $output/$sample/${sample}.rmdup.bam

    > bamToBed -i Sample.rmdup.bam > Sample.bed
  • h.
    Extract cytosine methylation from the BAM file using singleC_metLevel.hg19.pl in https://github.com/jasminexiao/niPGT. The DNA methylation level is calculated as the ratio of the number of reads with methylated C to that of total reads (methylated and unmethylated). Type command as follows:

> samtools view –h Sample.rmdup.bam | samtools view -uSb /dev/stdin | samtools mpileup -O -f hg19.genome.fa /dev/stdin > Sample.pileup

> perl singleC_metLevel.hg19.pl Sample .pileup > Sample.single5mC_tmp

> grep -v " chrM " Sample.single5mC_tmp > Sample.single5mC

Inferring copy number variations (CNVs)

Timing: 30 min

In this step, chromosome copy number variations are calculated.

• 29.Installing system requirements of Ginkgo toolkit.

Note: Ginkgo is a single-cell CNV analysis tool that can be downloaded from https://github.com/robertaboukhalil/ginkgo. Click in the green ‘Code’ tab for downloading the zip folder, which contains several directories and a ReadMe file. And then install the requirements according to the ReadMe file.

• 30.
  Inferring CNVs.

  • a.
    Enter the uploads directory in the ginkgo-master folder and create your own analysis directory. Type command as follows:

    > cd uploads

    > mkdir analyse
  • b.
    In the analysis catalog, prepare files consisting of config, list, sample.bed or sample.be.gz (e.g., PBAT_test_S1.bed.gz) and reference sample (e.g., XX_NG_ICM_E_4M.median.bed). Type command as follows:

    > cd analyse

    > rz config

    > rz XX_NG_ICM_E_4M.median.bed

    > rz list

    > rz PBAT_test_S1.bed.gz

    > sh ../../scripts/analyze.sh analyse

    Note: If there is no reference sample, the parameter can be modified as segMeth=0 in the config file.

  • c.
    Run analyze.sh to call CNV. Type command as follows:

> sh ../../scripts/analyze.sh analyse

Note: There is an example named PBAT_test_S1.bed.gz in https://github.com/jasminexiao/niPGT.

Deducing cellular components

Timing: 15 min

In this step, three cellular components of SECM cfDNA are inferred via DNA methylation.

Note: The cfDNA from the SECM of the ICSI-generated blastula has three main cellular components: the blastula, the cumulus cell, and the polar body. We have identified 769 cumulus-specific differentially methylated regions (C-DMRs) and 548 oocyte/polar body-specific DMRs (O-DMRs).¹ The average methylation levels of the C-DMRs of the blastula, the cumulus cell and the oocyte/polar body are 4%, 92% and 3%, respectively. The average methylation levels of the O-DMRs of the blastula, the cumulus cell, and the oocyte/polar body are 22%, 19% and 82%, respectively. The mathematical relationship between the methylation levels of each DMRs (DMRi, referring to C-DMRs or O-DMRs) in the SECM cfDNA and the proportional contribution of each component (component k, referring to the blastula, the cumulus cell, and the polar body) can be expressed by the formula MMi = ΣMCik × Pk × aik, in which MMi represents the methylation levels of DMRi in the SECM cfDNA, MCik represents the average methylation levels of DMRi in component k, Pk represents the proportional contribution of component k to the SECM cfDNA, and aik represents PCR amplification efficiency of DMRi in component k (Table 1). Therefore, there are three unknown Pk (P for the blastula, the cumulus cell and the polar body) in three formulas (the formulas for C-DMRs and O-DMRs, as well as the formula that the sum of three Pk is equal to 100%). By solving the three-variable linear equation, we can get the values for three Pk. The parameters of MCik and aik are listed in Table 1.

Table 1.

The parameters for deducing the proportions of the blastula, the cumulus cell, and the polar body

		Component k
		Blastula (MC; a)	Cumulus cell (MC; a)	Polar body (MC; a)
DMR i	C-DMRs	4%; 1	92%; 0.6	3%; 1
	O-DMRs	22%; 1	19%; 1	82%; 0.6

Note: “MC”: the average methylation level of DMR i in component k. “a”: correction factor for the potential bias of PCR amplification toward the unmethylated allele.¹⁹

• 31.Calculate the methylation level of C-DMRs for the SECM cfDNA.

> cat Sample.single5mC | grep -w 'CpG' | awk 'BEGIN{OFS="∖t"}{if($5>='3'){print $1,$2-1,$2,$5,$6}}' | bedtools sort -i > Sample.CpG_sorted.bed

> bedtools intersect -a C_DMRs.bed -b Sample.CpG_sorted.bed -wb -wa | awk '{OFS = "∖t"; print $1, $2, $3,$7,$8}' | bedtools groupby -g 1,2,3 -c 4,5 -o sum,sum | awk '{OFS = "∖t";print $1,$2,$3,$4,$5,$6=$5/$4}' > Sample_Cumulus.single5mC

• 32.Calculate the methylation levels of O-DMRs for the SECM cfDNA.

> bedtools intersect -a O_DMRs.bed -b Sample.CpG_sorted.bed -wb -wa | awk '{OFS = "∖t"; print $1, $2, $3,$7,$8}' | bedtools groupby -g 1,2,3 -c 4,5 -o sum,sum | awk '{OFS = "∖t";print $1,$2,$3,$4,$5,$6=$5/$4}' >Sample_polarbody.single5mC

• 33.Deducing the proportions of the blastula, the cumulus cell and the polar body within the SECM cfDNA.

> Rscript calculate_components_ratio.R

Note: All differentially methylated regions and scripts for calculating component ratios can be found in https://github.com/jasminexiao/niPGT.

Expected outcomes

Library quality control

After PCR amplification, the product is subjected to the “library quality control” step for verifying whether the library has corrected concentration and size distribution. For a typical SECM sample, the concentration is about 3 ng/μL or 75 ng in total. The size of the product ranges between 200 to 800 bp, with a peak locating between 300 to 450 bp (Figure 2). The primer dimers locate at 100 to 200 bp. In general, there should be no or little primer dimers as shown in Figure 2; if there are some primer dimers, the primer-dimer peak should be lower than 50% of the product peak, or an additional round of purification is required.

Figure 2.

The size distribution of a typical SECM-PBAT library

The X and Y axes show the fragment length (bp) and fluorescence units (FU), respectively.

Data quality control

For each sample, we sequence about 5-Gigabase (Gb, equal to 16.7 million 150-bp pair-end reads), which can get a mean of 12 million clean reads and 3.6 million uniquely mapped reads. This leads to a mean coverage of 5.3 million CpG sites (≥ 1×).

Since the SECM-PBAT method is highly sensitive, it can amplify potential trace DNA contamination from the reagent, the environment and the culture medium. In the data processing step, we have discarded R2 reads that have more than 3 unmethylated CHs, which should help eliminate DNAs that contaminate after bisulfite treatment. As a data quality control step, one should also routinely check the DNA methylation profiling across the transcriptional starting site (TSS), the gene body and the transcriptional end site (TES), which is expected to display a valley surrounding TSS and a plateau across the gene body (Figure 3).

Figure 3.

DNA methylation profile across the gene body

For three SECM samples with no (red), moderate (green) or severe (blue) cumulus DNA contamination respectively, the DNA methylation levels are profiled across the TSS, gene body, TTS, as well as upstream and downstream 15-kb flanking regions. Note the valley surrounding TSS and the plateau across the gene body.

The DNA methylation level of the blastula is about 24%.⁹ Thus, if the SECM cfDNA is solely derived from the blastula, it should have a similar methylation level. Since the DNA methylation level of the cumulus cell is about 71%, the methylation level will increase. This pattern can also be revealed on the DNA methylation profiling crossing the gene (Figure 3). The lambda DNA (unmethylated) is added to assess the bisulfite conversion rate, which is 0.99 on average in our data.

Simultaneous detection of chromosome aneuploidy and cellular origins for SECM cfDNA

The SECM-PBAT method can simultaneously detect chromosome aneuploidy and deduce cellular proportions of the SECM cfDNA. We have shown that SECM-PBAT gives similar copy number (CN) profiles as the single-cell whole genome amplification (scWGA) method including multiple annealing and looping-based amplification cycles (MALBAC).¹ Interestingly, the CN profiles of some SECM samples show relatively high chromosome-level variations though not detected as an aneuploidy by the software (Figure 4C). This is infrequently seen in our previous single-cell PBAT data.⁹ We speculate that this should be mainly due to the CNVs mosaicism of the embryo.

Figure 4.

Simultaneous detection of chromosome aneuploidy and cellular origins of SECM cfDNA

(A–D) The CN profiles and the proportional contributions of the blastula, the cumulus cell and the polar body of four SECM samples with no (A), moderate (B and C) and severe (D) maternal DNA contamination (the sum proportions of the cumulus cell and the polar body) are shown. In the CN profiles (left), each dot represents a 1-Megabase (Mb) window, with a red line representing copy number duplications, a blue line representing copy number deletions and a black line representing normal copy number. The CN results of the SECM samples and the diagnostic results of the corresponding TE biopsies are shown on the top, with the proportional contributions of the maternal DNAs including the cumulus cell (C) and the polar body (Pb) contamination. The pie charts (right) show the proportions of each component including the blastula (yellow), the cumulus cell (blue) and the polar body (red).

The SECM-PBAT method also deduces the proportional contributions of the blastula, the cumulus cell and the polar body based on the methylation levels of C-DMRs and O-DMRs in the SECM cfDNA. The proportions of the cumulus cell and the polar body make up maternal DNA contamination. By integrating the maternal contamination ratio and chromosome aneuploidy information, we have shown that it shows promising to increase the diagnostic accuracy of niPGT-A.¹ Figure 4 shows the chromosome aneuploidy results of four SECM cfDNA samples with different proportions of the blastula, the cumulus cell and the polar body.

Limitations

One limitation is that the present studies have only examined the ICSI embryos.¹ We have found that natural inseminated embryos have additional contamination derived from the sperm, and are identifying sperm-specific DMRs for simultaneously deducing the proportional contribution of this component. These data are still unpublished and will be reported in the future.

In addition, several single-cell whole genome DNA methylation sequencing methods have been reported, and these methods should also be suitable for SECM cfDNA methylation analysis.²⁰

Troubleshooting

Problem 1

In step 5, the efficiency of bisulfite conversion of unmethylated C into T is too low. The conversion rate as assessed by the spike-in lambda DNA should be more than 0.95.

Potential solution

CT Conversion Reagents expire or fail due to lack of light protection during storage, or are not sufficiently dissolved before use. The CT Conversion Reagents need to be strictly protected from light, used within the validity period, and vortexed until fully dissolved and used before expiration.

Problem 2

In step 24, the concentration of purified DNA from PCR is less than 0.2 ng/μL.

Potential solution

Too much DNA loss before PCR amplification, so use LoBind tubes during the whole procedure. AMpure beads should not be excessively dry, otherwise the amount of DNA will be reduced, thus reducing the efficiency of PCR amplification. It is also possible that MagBinding Beads are not kept in suspension during the purification step after CT conversion. So during the purification process, pay attention to the status of the MagBinding Beads to recover the maximum amount of DNA.

Problem 3

In step 25, the primer-dimer is excessive.

Potential solution

Perform an additional round of AMPure XP bead purification to remove the excessive primer-dimer. Ensure that the PCR primer concentration and the PCR cycle number are appropriate.

Problem 4

In steps 24 and 25, positive library products are amplified from the nuclease-free water as a negative control.

Potential solution

Include the nuclease-free water as a negative control for each set of experiments. Positive library products indicate possible contamination during steps 1–22. Use pipette tips with a filter and perform the library construction experiment in a designated workstation which receives regular ultraviolet irradiation. Clean the operation area and the pipettor regularly. Replace with a new batch of reagents when necessary. If the quantity of the library product amplified from the nuclease-free water is less than 50% of those from the SECM samples, process with sequencing. Sequencing the library product amplified from the nuclease-free water to check the source of contamination. Discarding R2 reads that have more than 3 unmethylated CHs helps reduce genomic DNA contamination in steps after bisulfite treatment.

Problem 5

In steps 24 and 25, positive library product are amplified from the culture medium as a negative control.

Potential solution

Include the culture medium as another negative control for each set of experiments. A higher quantity of library products amplified from the culture medium than from the nuclease-free water amplification indicates possible contamination in the culture medium. Use batches of culture medium with no contamination. Perform IVF, embryo culture and SECM collection in a sterile IVF Workstation. Use pipette tips with a filter and clean the pipettor regularly. If the quantity of the library product amplified from the culture medium is less than 50% of those from the SECM samples, process with sequencing. Sequencing the library product amplified from the culture medium to check the source of contamination.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Lu Wen (wenlu@pku.edu.cn).

Materials availability

This study did not generate new unique reagents.

Data and code availability

Original data have been deposited to the National Genomics Data Center of the China National Center for Bioinformation (https://ngdc.cncb.ac.cn/gsa-human/): HRA000332. Codes for relevant analysis are available at https://github.com/jasminexiao/niPGT and https://zenodo.org/record/7726437.