Quantitation of DNA Methylation by Quantitative Multiplex Methylation-Specific PCR (QM-MSP) Assay

Abstract

The defining feature of the Quantitative Multiplex Methylation-Specific PCR (QM-MSP) method to sensitively quantify DNA methylation is the two-step PCR approach for a multiplexed analysis of a panel of up to 12 genes in clinical samples with minimal quantities of DNA. In the first step, for up to 12 genes tested, one pair of gene-specific primers (forward and reverse) amplifies the methylated and unmethylated copies of the same gene simultaneously and in multiplex, in one PCR reaction. This methylation-independent amplification step produces amplicons of up to 10⁹ copies per μL after 36 cycles of PCR. In the second step, the amplicons of the first reaction (STEP 1) are quantified with a standard curve using real-time PCR and two independent fluorophores to detect methylated/unmethylated DNA of each gene in the same well (e.g., 6FAM and VIC). One methylated copy is detectable in 100,000 reference gene copies. Methylation is reported on a continuous scale. For the gene panel, the highest level of normal DNA methylation above which a sample would be called positive is derived by using Receiver Operating Characteristic (ROC), maximizing assay specificity and sensitivity to distinguish between normal/benign versus tumor DNA. QM-MSP can be applied to clinical samples of fresh or fixed ductal cells, ductal fluid, nipple fluid, fine needle aspirates, core biopsies, and tumor tissue sections.

1. Introduction

Quantitative multiplex methylation-specific PCR (QM-MSP) [1, 2] is a highly sensitive, specific, and quantitative methylation assay used by our laboratory and others [1-10] (Fig. 1a). It combines the principles of conventional gel-based MSP [11] (see also Chapter 23), quantitative real-time MSP (qMSP) [12, 13] (see also Chapters 23 and 25) and multiplexed gel-based MSP [14] into one format developed to enable quantification of methylated gene panels in clinical samples with limited DNA quantities available (pg-ng; 50–1000 cells). The assay is easily performed on fresh or fixed cytological samples including ductal lavage/ductoscopy fluids and cells, nipple fluids, and fine needle aspirates as well as tissues and core biopsies. The QM-MSP method has an analytical sensitivity of 1 methylated copy in 100,000 unmethylated copies [2, 3, 10], nearly tenfold higher than qMSP, and 100-fold higher than gel-based MSP techniques. With more than double the clinical sensitivity for detection of cancer compared to cytology alone [2, 10], a panel of genes can be co-amplified and quantified on a continuous scale, discriminating normal or benign versus cancer samples using an ROC-derived threshold.

Fig. 1.

Schema of QM-MSP. A two-step PCR is performed where the first PCR reaction (STEP 1) contains one pair of external primers per gene (forward and reverse) that co-amplifies DNA from the methylated gene of interest (TARGETgene; M) and the gene-specific unmethylated allele (U). In the second PCR reaction (STEP 2), amplicons of step 1 are assayed by absolute quantitative real-time PCR with specific sets of primers ( forward and reverse) and hydrolysis probes (in two colors) recognizing methylated TARGETgene. Figure adapted by permission from the American Association for Cancer Research: Fackler et al., Quantitative multiplex methylation-specific PCR assay for the detection of promoter hypermethylation in multiple genes in breast cancer. Cancer Res. 2004; 64:4442–52 [2]

In the protocols described in this chapter, we present methods for QM-MSP as well as instructions for preparation and handling of genomic DNA extracted from fresh cells or formalin-fixed, paraffin-embedded tissue, samples obtained from fine needle aspirate, ductoscopy, ductal lavage, core needle biopsy, excisional biopsy tissue and tumor. We have included all the different steps necessary for the processing of patient samples, sodium bisulfite treatment of DNA, construction and calibration of standard curves, primer design, setup and execution of multiplex PCR, and setup and execution of real-time PCR.

2. Materials

All chemicals and solutions must be handled with PCR safe techniques. Unless otherwise specified (see Note 1), stock chemicals are stable indefinitely at room temperature.

2.1. Buffers

1. 1 M NH₄SO₄, prepare fresh each time.

2. 1 M Tris–HCl, pH 8.0.

3. 2 M Tris–HCl, pH 8.8.

4. 1 M MgCl₂·6H₂O.

5. TE: 10 mM Tris–HCl, pH 8.0, 1 mM EDTA, pH 8.5.

6. 5 M NaCl.

7. 500 mM EDTA, pH 8.0.

8. DTT: 1 M dithiothreitol.

9. 10% SDS.

10. β-mercaptoethanol (Sigma; 14.3 M), keep stock at −20 °C, replace every couple of months.

11. DMSO, PCR grade.

12. Isopropanol.

13. Ethanol (100%, 70%, 50%).

14. Xylene.

15. ddH₂O.

16. Proteinase K (20 mg/mL).

17. RNase A-, DNase-, and protease-free (10 mg/mL).

18. TNES: 10 mM Tris, pH 8.0, 150 mM NaCl, 2 mM EDTA, 0.5% SDS.

19. TNES/SS: 1 mL TNES, 5 μL Salmon sperm DNA (10 mg/mL).

20. TNES/SS/PK: 900 μL TNES/SS, 100 μL Proteinase K (20 mg/mL).

21. 10× MSP Buffer: 166 mM NH₄SO₄, 670 mM Tris–HCl, pH 8.8, 67 mM MgCl₂, 100 mM β-mercaptoethanol, 1% DMSO (see Note 2).

22. 1× Dilution Buffer: 100 μL 10× MSP buffer, 5 μL salmon sperm DNA 10 mg/mL, 5 μL tRNA, 10 mg/mL, 890 μL ddH₂O).

23. DNA Hydrating Solution (LowTE): 1 mM Tris–HCl, pH 8.0, 0.1 mM EDTA.

24. PBS (1×): 10 mmol Na₂HPO₄, 1.76 mmol KH₂PO₄, 137 mmol NaCl, 2.7 mmol KCl without calcium or magnesium, pH 7.4 (e.g., Corning, 21-040-CV).

2.2. Controls

1. M.SssI methylase (New England Biolabs).

2. PUREGENE DNA Purification Kit (Qiagen) or similar.

3. NanoDrop (Thermo Fisher Scientific).

2.3. Sample Preparation

1. PAP jar.

2. Temperature-controlled oven.

2.4. PCR Reagents

1. Forward and reverse gene-specific external primers.

2. Gene-specific forward and reverse internal primers and probes (FAM- or VIC-labeled fluorophore and TAMRA as a quencher).

3. Platinum Taq polymerase (Thermo Fisher Scientific).

4. dNTPs.

5. ROX 50× (Thermo Fisher Scientific).

6. tRNA (10 mg/mL).

7. RAMP Taq 5 U/μL (Denville Scientific).

8. 7500 Real-time PCR System (Thermo Fisher Scientific) or similar.

3. Methods

3.1. Primer Design

Software is available from several commercial sources that can be used to design primers and probes for methylation-specific PCR. However, we recommend the manual design of primers/probes. By convention, forward primers/probes use sense sequences and reverse primers use the reverse complement of the sense sequence. Primer sequences are written in the 5′-3′ orientation. For each gene, primers for a methylated TARGETgene (M), and a REFERENCEgene [unmethylated TARGETgene (U) is used as reference for QM-MSP] are designed. For example, M TARGET_RASSF1, REFERENCE_RASSF1 (U TARGET_RASSF1 for QM-MSP).

1. Find an area that is rich in CGs overall. When studying the role of gene promoter hypermethylation in silencing gene expression, usually choose an area to situate primers/probes near the translational start site of the gene. When studying genes discovered by molecular methods such as methylome array (see also Chapter 16), choose a location for QM-MSP primers/probes within 100 bp of the array probe, if possible (see Note 3 for an example analyzing AKR1B1).

2. Copy/paste the genomic RefSeq nucleotide sense strand sequence for the CpG region of interest plus several hundred upstream and downstream bases, into a word processing document. Make all bases lower case. Find/replace all CG dinucleotides with bold upper case letters. Find/replace all lower case “c” with “t.” This is now the predicted sequence of bisulfite-converted methylated DNA. Save.

3. Copy/paste the predicted sequence of bisulfite-converted methylated DNA to a new section of the document and within the pasted text find/replace all bold C residues with bold T residues. This is now the predicted sequence of bisulfite-converted unmethylated DNA. Save.

4. Design of the Multiplex methylation-independent external primers for each gene for STEP 1: Primers specifically co-amplify bisulfite converted DNA template for both the TARGETgene as well as the REFERENCEgene. Up to 12 TARGET/REFERENCE genes can be amplified in one tube. Select sequences for external primers based on the following:

   1. Position each external primer sequence in a CpG poor stretch of bases flanking or within the CG rich area. Ideally, external primers should be positioned to amplify 150–250 bp DNA and be 20–25 bp each in length, with a predicted melting point (Tm) of 56–58 °C (based on the C or G = 4 °C, A, or T = 2 °C rule).
   2. Several internal stand-alone C residues should be present.
   3. Avoid repetitive nucleotides of >4 bases. Try to have the last base or two at the 3′ end of the primers finish with a C or G. Between the binding sites for forward and reverse primers there should be a sufficient number of bases, including CpG dinucleotides, to accommodate the internal region of the next PCR step.
   4. Primers must not be complementary to each other or contain palindromes.
   5. External primers should amplify the target sequences independent of the methylation status and therefore should not contain CpG dinucleotides. If this cannot be avoided they should be synthesized to have equimolar degenerate nucleotides for cytosine at the CG position, designated as “Y” (C + T) in the forward primer and as “R” (G + A) in the reverse primer rendering the amplification methylation-independent (see Note 4).
5. Design of Real-time STEP 2 Methylation-specific Internal Primers/Probes that co-amplify diluted amplicons from STEP 1. Select sequences for the methylation-specific inner primers/probes based on the following criteria:

   1. Methylated target gene set: To design the methylated TARGETgene primer/probe set, begin with the sequence predicted for methylated bisulfite converted DNA and select a forward and reverse primer pair positioned to amplify 80–100 bp DNA within the region amplified by the external primers in step 4 above. Internal primers should be 20–25 bp each, have a predicted Tm = 64-66 °C, several independent stand-alone C residues, 2–3 CpG dinucleotides per primer, and end with a 3′ C or G if possible.
   2. Avoid repetitive stretches of the same nucleotide, as well as CGCG regions and palindromes. Primers must not be complementary to each other (see Note 4) or the hydrolysis probe.
   3. The hydrolysis probe is usually designed in the reverse orientation, should be 25–30 bp, have a predicted Tm = 74–76 °C, several independent stand-alone C residues, 3–4 CpG dinucleotides, and must not have a 5′ G residue in the first or second position. Ideally the probe will terminate with a 3′ C or G. The 5′ end should be linked to an intense fluorophore such as 6-FAM, although others such as VIC are adequate, with the 3′ end linked to a quencher such as TAMRA (see Note 5).
   4. Unmethylated target gene set: To design the unmethylated REFERENCEgene primer/probe set use the sequence predicted for unmethylated bisulfite converted DNA and perform the same steps as above. Try to match the Tm of the unmethylated primers to that of the methylated primers. Situate the forward primer, reverse primer and probe overlapping the sites for methylated primers/probe. Unmethylated primers/probe should be 2–3 nucleotides longer than the methylated set to compensate for the conversion of CG to TG and thus the change in predicted Tm.

3.2. Universally Methylated DNA Stock (M.SssI CpG Methylase Reaction)

Fully methylated control DNA is used as “M” control in QM-MSP STEP 1, and subsequently in STEP 2, the M amplicons are used as an M control. Another use is to establish the gene-specific M standard curve for real-time PCR. There are several commercial sources of universally methylated DNA. However to prepare it yourself, purify genomic methylated cancer cell line DNA (e.g., MDA-MB-453), and then methylate this DNA by treating the DNA with the CpG methylase enzyme M.SssI. SssI enzyme methylates cytosine residues in the sequence CpG. The protocol described uses the reagents supplied by New England Biolabs, but the M.SssI enzyme is commercially available from several sources.

Definition: 1 U of SssI methylates 1 μg of DNA (lambda) in 1 h at 37 °C in the presence of 160 μM S-adenosyl methionine (SAM). To drive the forward reaction, high amounts of fresh SAM must be used and SAM/SssI must be replenished during the incubation with the methylase.

1. Assemble the reaction mix by adding 2 μg of DNA, 5 μL NEB 2 buffer (500 mM NaCl, 100 mM Tris–HCl, 100 mM MgCl₂, 10 mM DTT, pH 7.9 , or similar buffer), 1 μL SAM (supplied stock: 32 mM stored at −80 °C, freshly thawed,) and 5 μL SssI enzyme (4 U/μL).

2. After 3 h at 37 °C add another 50 μL of freshly prepared reaction mix (no DNA), continue to incubate an additional 3 h 37°C.

3. Heat inactivate at 95 °C for 5 min.

4. Perform the sodium bisulfite DNA conversion reaction directly after this step or store sample frozen at −80 °C. There is no need to clean up the DNA prior to the sodium bisulfite reaction step.

3.3. Unmethylated DNA Stock

Human sperm DNA, leukocyte DNA or other cell source lacking methylation at the locus of interest may be used as a source of control unmethylated DNA. Unmethylated control DNA is used as “U” control in the STEP 1 multiplex PCR, and subsequently in STEP 2 the U amplicons are used as an U control. Another use is make the gene-specific U standard curve for real-time PCR. For leukocytes, prepare genomic DNA using routine phenol/chloroform extraction or commercially available kits.

Human sperm DNA (HSD) can be purified using the PUREGENE DNA Purification Kit:

1. Dilute 5 mL fresh seminal fluid to 25 mL with TE and place on ice.

2. Centrifuge the specimen at 2100 × g for 15 min at 4 °C.

3. Discard the supernatant and vortex the cellular pellet rigorously for 1 min, then resuspend in 6 mL of PureGene Cell Lysis Solution, 240 μL 1 M dithiothreitol, and 30 μL Proteinase K (20 mg/mL).

4. Mix the sample by inverting 25 times and incubated at 55 °C overnight.

5. Add 30 μL RNase A solution (10 mg/mL, provided in the kit), to the cell lysate and mix the sample by inverting another 25 times, and then incubate at 37 °C for 60 min.

6. Cool the sample to room temperature; add 2 mL of Protein Precipitation Solution and vortex the lysate vigorously for 20 s.

7. Place the sample in an ice bath for 5 min, and then centrifuge at 2000 × g for 10 min.

8. Transfer the supernatant (~8 mL) to a 50 mL conical tube.

9. Precipitate the DNA in the supernatant by adding 2 mL 10 M ammonium acetate, and 25 mL absolute ethanol, then invert 50 times, and centrifuge at 3000 × g for 15 min. Wash the DNA pellet in 80% ethanol. Air dry, and then rehydrate the pellet in 400 μL DNA Hydrating Solution. Incubate at 65 °C 1 h until completely dissolved. Store at 4 °C.

3.4. Standard Curve Stocks for STEP 2, Real-Time PCR

For each gene, master PCR stocks of (1) methylated TARGETgene (“M”) and (2) REFERENCEgene (U) for the real-time standard curves, are made ahead and frozen (−80 °C) indefinitely. To make the working standard curve stock, small aliquots of target and reference DNAs are mixed in equimolar ratios and these “Standard Curve” stocks are kept frozen as a gene-specific master stock. Note that each gene has its own standard curve stock where M and U are in equimolar amounts. Because the stock is too concentrated to use directly, a dilution of 1:10 is prepared from the master standard curve stock. From the 1:10 dilution a standard curve with dynamic range of 10⁻²–10⁻⁸ is prepared each time the real-time run is performed. The 1:10 dilution is refrozen and stored at −80 °C.

1. Prepare single-gene reaction mixes using the recipe shown in Subheading 3.7 (STEP 1 multiplex PCR recipe) using 20–40 ng DNA of either the universally methylated (M) or the unmethylated (U) DNA and 2 μL each of external forward/reverse primers (stock is 100 ng/μL).

2. Perform a PCR reaction in a 500 μL PCR microfuge tube using a final volume of 50 μL volume. Thermocycler settings are 95 °C for 5 min, followed by 36 cycles of 95 °C for 30 s, 56 °C for 45 s, and 72 °C for 45 s, with a final extension of 72 °C for 7 min, and a 4 °C hold.

3. Visualize 3 μL of the PCR product using 3% agarose gel electrophoresis. Products should be single bands of the expected size. Store the stocks undiluted at −80 °C.

4. Final “Standard Curve Stocks” have equimolar M TARGETgene and REFERENCE gene (U) mixed together in one tube for a single gene. To prepare this stock for a single gene, begin by combining equal volumes of the M only and U only stock PCR reactions above (e.g., 3 μL each in one 500 μL tube, only 1 gene per tube).

5. Make serial 1:10 dilutions (e.g., 3 μL DNA + 27 μL 1× dilution buffer) to prepare seven dilutions ranging from 10⁻² to 10⁻⁸ of the M/U mix.

6. Assay 4 μL of each dilution per well by real-time PCR (M target and reference DNAs assayed in the same well) and determine the cycle threshold (Ct) at each dilution for M and U.

7. Calculate the average ΔCt between M and U at each point over the entire curve (10⁻²–10⁻⁸). Based on this ΔCt, calculate the fold difference in copies: for example if average ΔCt = 1.5 cycles, then 2^1.5 = 2.8-fold difference on average over the entire curve.

8. From aliquots of the master PCR stocks, remix the M only and U only controls to make an equimolar mixture of M and U (the real-time PCR amplification plots will overlap), taking into account the fold difference in copies and adding more of the lowest concentration control, i.e., 3 μL of M + 8.4 μL (3 μL × 2.8-fold) of U, if U was at the lowest concentration.

9. Retest the new mix to ensure that M and U curves are now overlapping (≥0.5 cycles difference). Store the Master PCR stocks (M or U) at −80 °C (stable indefinitely).

10. Dilute the M/U stock 1:10 in 1× dilution buffer and this will be the working M/U curve stock. Freeze these stocks at −80 °C indefinitely (see Note 6).

11. In step 6 (Subheading 3.8) serial 1:10 dilution curves will be made from the working M/U curve stock for a dynamic range of 10⁻²–10⁻⁸ for every real-time PCR. If the working M/U curve stock begins to drift after several freeze/thaw cycles, remake the 1:10 dilution from the M/U master standard curve stock.

3.5. Samples

3.5.1. Fresh Cells

As few as 50–100 cells can be assayed by QM-MSP, but 1000–20,000 cells are preferred.

1. Centrifuge freshly isolated cells in phosphate buffered saline (PBS), pH 7.4.

2. Discard the supernatant, transfer cells to a 500 μL microcentrifuge tube, briefly centrifuge to pellet the cells, discard the supernatant and resuspend the cells in 30 μL TNES/SS/PK (for cell numbers greater than 20,000 use up to 100 μL TNES/SS/PK).

3. Incubate the sample at 56 °C overnight, add 1/10th volume of additional proteinase K (20 mg/mL), continue to incubate for 4 h, heat inactivate at 95 °C for 10 min and store the sample at 4 °C.

3.5.2. Frozen Tissue Embedded in OCT Compound

1. Thaw frozen tissue mounted on a glass slide and dissolve the embedding compound in 70% ethanol for 1 min, then 50% ethanol for 5 min.

2. Wash tissue by dipping the slide up and down gently until all the OCT rinses off.

3. Then dip the slide five times in 70% ethanol, drain and air dry the sample.

4. Harvest, at a minimum, a 4–5 mm² area of tissue.

5. With a fresh single-edge razor or pipet tip, scrape the tissue into a microfuge tube containing 27 μL TNES/SS (see NoteS 7 and 8).

6. Add 3 μL proteinase K (20 mg/mL) and continue with the proteinase K digestion as described in Subheading 3.5.1.

7. For slices of tissue in OCT not on slides place the cryosection/s into a 1.5 mL microfuge tube and incubate in 1 mL 50% ethanol for 5 min, pulse vortex 1 min, centrifuge briefly to pellet the tissue, remove the supernatant.

8. Repeat with 70% ethanol. Carefully remove the supernatant and pellet the sample.

9. Add 30 μL TNES/SS/PK (for a large tissue section use up to 100 μL of TNES/SS/PK) and digest the tissue as described in Subheading 3.5.1 (see Note 7).

3.5.3. Formalin-Fixed, Paraffin-Embedded (FFPE) Tissue Sections

QM-MSP requires at least 4–5 mm² tissue depending on the quality of the DNA, which is routinely fragmented by formalin and further damaged by prolonged fixation.

1. Paraffin-embedded tissue on glass slides should be incubated in 40 mL fresh xylene for 10 min at room temperature in a disposable PAP jar, and the slide dipped up and down ten times.

2. Repeat once with new xylene.

3. Air-dry the slide in a chemical hood.

4. With a fresh sterile flat single-edge razor blade, scrape the tissue into a 500 μL microcentrifuge tube containing 30 μL TNES/SS/PK (see Notes 7 and 8). For a large tissue section, use up to 100 μL of TNES/SS/PK.

5. Process the sample overnight as described for proteinase K digestion in Subheading 3.5.1.

3.5.4. FFPE Tissue Plugs/Cores

1. Deparaffinize FFPE cores or slices from tissue blocks with xylene in a 1.5 mL microfuge tube.

2. Incubate the tissue in 1 mL fresh xylene for 10 min, vortex for 1 min, centrifuge it for 5 min at full speed and remove the supernatant.

3. Repeat twice more with 1 mL fresh xylene, drain.

4. Add 1 mL of absolute ethanol to wash off the xylene from the tissue, centrifuge and drain well.

5. Add 100 μL TNES/SS/PK (see Note 7).

6. Digest the sample as described in Subheading 3.5.1.

7. Use 5–10 μL of the lysate for the QM-MSP assay and store the rest of the lysate at 4 °C or colder.

3.5.5. Stained Cytology Samples

PAP stained cellular samples such as those prepared using a cytocentrifuge or ThinPrep device, mounted with a coverslip can be assayed (see Note 9):

1. Dissolve the mounting media holding the coverslip in place by incubating in xylene for 2–3 days or longer until the coverslip falls off on its own. Do not force it off or some cells may adhere.

2. Rinse the slide by dipping in fresh xylene and air-dry.

3. Wet the section with 20–40 μL of TNES. With a clean pipette tip or single edge razor blade, scrape cells into a 500 μL microcentrifuge tube taking care to rinse the cells repeatedly with TNES/SS (see Note 7).

4. Briefly centrifuge the sample to pellet the cells.

5. Remove all TNES/SS and resuspend the cell pellet in 30 μL TNES/SS/PK. Proteinase K-treat the sample as described in Subheading 3.5.1.

3.6. Protocol for Sodium Bisulfite Conversion of the DNA Template

Sodium bisulfite chemically converts unmethylated cytosine residues in DNA to uracil, which is read as thymidine during PCR. When cytosine is protected by DNA methylation (i.e., contains a 5′ methyl group, ^mCG) it resistant to sodium bisulfite and there will be no nucleotide sequence change of the ^mCG after incubation with sodium bisulfite. This is the basis of the methylation specific PCR (MSP) method described by Herman et al. [11]. Primers that are predicted to differentially hybridize to either methylated or unmethylated DNA after the sodium bisulfite reaction are referred to as methylated (M) or unmethylated (U) methylation-specific PCR (MSP) primers, respectively. Many commercial sodium bisulfite DNA conversion kits are available, including the EZ DNA Methylation Kit (D5001; Zymo Research) described in this subheading. This protocol is a modified version of the manufacturer’s instructions.

1. Denature the double stranded DNA: Mix DNA (up to 3 μg purified DNA or 40 μL of lysate in TNES/SS/PK) + ddH₂O to make a final volume of 42.5 μL in a 500 μL microcentrifuge tube.

2. On ice, add 7.5 μL M-dilution buffer (supplied with the kit). Heat samples at 42 °C 15 min to denature the DNA (see Note 10). Chill samples on ice.

3. Perform the bisulfite conversion: During the incubation in step 1, prepare the CT reagent (supplied) and use it immediately after the 42 °C incubation is finished: On a per vial basis (small vial sufficient for 10.5 samples), add 750 μL ddH₂O + 185 μL M-dilution buffer to the vial containing 567 mg of CT reagent. On a per sample basis, combine 71.4 μL water + 17.6 μL M-dilution buffer and 54 mg of conversion reagent. Dissolve the CT conversion reagent by rotating it in the dark at room temp for 10 min. When the denaturing step is completed (step 2), add to each sample 97.5 μL CT conversion reagent. Pulse vortex. Centrifuge 10 s.

4. Incubate the sample overnight in a thermocycler: 16 cycles of 95 °C 30 s, 50 °C 1 h (see Notes 11 and 12). The final volume is 150 μL. Hold the sample at 4 °C until the DNA clean-up step.

5. DNA clean-up: Using a 1 mL pipet, add four volumes of M-Binding Buffer (supplied) into a Zymo-Spin IC column seated in a 2 mL collection tube (supplied) and then add the converted DNA (i.e., 150 μL bisulfite reaction and 600 μL M-Binding Buffer).

6. Use the pipet tip to gently mix the sample with the binding buffer pipetting up and down 8–10 times (see Note 13).

7. Centrifuge at full speed 30 s at room temperature and transfer the column to a clean collection tube (see Note 10).

8. Add 100 μL M-Wash Buffer (supplied) to the column and centrifuge full speed for 30 s (see Note 11).

9. Add 200 μL M-Desulfonation Buffer (supplied) to the column, incubate 15 min at room temperature, and centrifuge full speed for 30 s (see Note 11).

10. Immediately neutralize the column by adding 200 μL M-Wash Buffer and centrifuging it at full speed for 30 s [it is important to wash away the Desulfonation buffer promptly, (see Note 11)].

11. Change to a new collection tube. Add 200 μL M-Wash Buffer to the column and centrifuge full speed for 1 min.

12. Transfer the column to a new collection tube.

13. Add 12 μL of warm water (70 °C).

14. Incubate the column with the water for 5–10 min at room temperature and centrifuge the column 1 min at full speed to recover the DNA. Transfer tubes to ice.

15. Use immediately for STEP 1 Multiplex PCR (20 ng/reaction is ideal) or store at −80 °C. For small amounts of DNA use the entire sample (10 μL) for the STEP 1 reaction.

3.7. Protocol for STEP 1: Multiplex PCR Reaction for ≤12 Genes

1. Perform the STEP 1 multiplex reaction in a 500 μL PCR microfuge tube using a final 50 μL reaction volume (Table 1) containing 16.6 mM NH₄SO₄, 67 mM Tris pH 8.8, 6.7 mM MgCl₂, 1.25 mM of each dNTP 10 mM β-mercaptoethanol, 0.1% DMSO, 2 ng/μL per external primer and 10 U/50 μL Platinum Taq polymerase (see Note 14). Thermocycler settings are 95 °C for 5 min, followed by 36 cycles of 95 °C for 30 s, 56 °C for 45 s, and 72 °C for 45 s, with a final extension of 72 °C for 7 min.

2. After completion of PCR, the reaction mix is diluted 1:5 by adding 200 μL 1× Dilution Buffer before performing the quantitative real-time PCR assay (STEP 2).

3. Include controls such as 100% methylated TARGETgene DNA (e. g., universally methylated DNA), unmethylated DNA (e.g., leukocyte or human sperm DNA), and water (no template control).

4. Optimal DNA input concentration is 20 ng, however up to 1000-fold less input DNA can be amplified if the sample has good integrity of the DNA (not formalin fixed).

5. The reaction mix can be stored frozen at −20 °C indefinitely. Prior to real-time PCR (STEP 2) each sample is further diluted 1:10² and/or 1:10⁴ (step 1 in Subheading 3.8).

Table 1.

Preparation of Multiplex STEP 1 Reaction Mix for One Sample, 12 genes

Multiplex STEP 1 reaction mix	1×
	Master mix (μL)	Final concentration
10× MSP	5.0	1×
25 mM dNTP	2.5	1.25 mM
Platinum Taq (5 U/μL)	2.0	0.2 U/μL
Water	6.5
Subtotal μL	16.0
For each of 12 genes:
External primer 100 ng/μl forward	1.0 × 12 genes	2 ng/μL per gene
External primer 100 ng/μl reverse	1.0 × 12 genes	2 ng/μL per gene
Subtotal μL	24.0
DNA (bisulfite-treated) ≤ 40 ng total	10.0
Total μL	50

3.8. Protocol for STEP 2: Quantitative Real-Time MSP (Two-Color qMSP)

The amplicons produced in STEP 1 are assayed in the same well in duplex using methylated (M) target and reference (U) sets of primers situated internally with respect to the multiplex primers. For each gene, the methylated target and reference U amplicons are tested with a unique set of primers (forward and reverse) and hydrolysis probes specific for M gene target or U reference gene.

1. Preparation of Sample & Control Dilution for STEP 2:

   1. Stocks of STEP 1 multiplexed patient DNA samples and U or M controls are stored frozen at −20 °C at a dilution of 1:5 in 1× Dilution Buffer.
   2. Thaw each sample at room temperature, and prepare a working stock by further diluting the sample/control 1:10⁻² and/or 1:10⁻⁴. These working stocks can be preassembled in a 96-well “template” plate, where the samples are organized in the same layout as in the STEP 2 real-time PCR plate.
   3. Apply an adhesive foil cover to the template plate, freeze at −20 °C.
   4. Before using, thaw and centrifuge the template plate (115 × g for 5 min) then remove the foil adhesive cover carefully in order to prevent cross-contamination of the wells. Repeated freeze/thaws will not affect the samples/controls.
2. Preparation of a master stock of primers/probes for both the target and reference sets: Keep the probes wrapped in foil or otherwise in the dark.

   1. Thaw individual primers and probes, vortex, combine and place on ice.
   2. Make enough for an entire run (e.g., 99× plus extra = 110× for one full plate) and label this tube “primers/probes”. For example, for the 110× stock combine 352 μL of each of the four primers with 4.4 μL of each of the two probes. This stock is Tube A. Refreeze master stocks.

3. Preparation of Master Reaction Mix: On ice, prepare sufficient amount of reaction master mix for one entire plate plus extra (e.g., 110× per plate). Following the recipe shown in Table 2 below, combine 10× MSP, dNTP, ROX, tRNA, water and Taq polymerase sufficient for 110× (see Note 14). For example, standards require 3× reaction mix for each of the seven dilutions = 21×, a full plate of 76 samples requires 76×, plus 6× for six controls = 106×, including extra = 110×. Controls include M only, U only, multiplex water, real-time PCR water, and known samples. Set aside.

4. For sample/control wells: Label a 1.5 mL centrifuge tube “All” and transfer into it the amount of primers/probes prepared in step 2 needed for wells B3–12, and C1 through H12 (Fig. 2).

5. Add the appropriate amount of reaction mix prepared in step 3 of this subheading. Example for 85×: Combine 1095 μL of primer/probe + 265 μL of reaction mix. Pulse vortex, briefly centrifuge. Keep on ice. This stock is Tube B.

6. The stocks for Standard Curves for Absolute Quantitation were prepared in Subheadings 3.2-3.4 and are stored at −80 °C. They were prepared separately for each gene and they consist of equimolar amounts of fully methylated target DNA and fully unmethylated DNA. A 1:10 M/U or M gene curve stock is kept as the working curve stock at −80°C and is thawed to prepare dilutions of the standard curve for STEP 2 real-time PCR.

   1. Set up a series of seven 500 μL tubes and place 27 μL of water containing salmon sperm DNA (50 μg/μL) into each tube. Perform a series of serial 1:10 dilutions of the Working Standard Curve stock by transferring 3 μL DNA to the first tube, mixing well, and transferring 3 μL of that DNA to the next tube using a new pipette tip, continuing in this way for all seven tubes. This will make 10⁻², 10⁻³, 10⁻⁴, 10⁻⁵, 10⁻⁶, 10⁻⁷ and 10⁻⁸ dilutions. Vortex, keep these tubes on ice.
   2. Set up another set of seven 500 μL tubes and place 9.36 μL reaction mix in each tube. Beginning with the first DNA dilution, transfer 12 μL of each dilution to respective tubes of reaction mix, cap, and vortex and keep these tubes on ice. These tubes will be used directly for STEP 2. Label these Tubes C 10⁻²–C 10⁻⁸. Refreeze the working M/U or M 1:10 curve stock.
7. Verify the following tubes and samples are ready, keep on ice:

   1. Sample DNAs and controls diluted 10⁻² and/or 10⁻⁴ (step 1).
   2. Tube A: “Primers/probes” for target and reference sets combined (step 2).
   3. Master Reaction Mix (step 3).
   4. Tube B: “All”, primers/probes (from Tube A) + reaction mix in sufficient amount for all samples and controls (step 5).
   5. Tube C tubes 10⁻², 10⁻³, 10⁻⁴, 10⁻⁵, 10⁻⁶, 10⁻⁷, 10⁻⁸ “10⁻²–10^−8”: Working mix of 3× volume of curve DNA + 3× volume of reaction mix per tube (step 6).

8. Assemble the 96-well plate on ice using plates recommended by the manufacturer of the real-time PCR instrument.

9. Pipet 12.88 μL of Tube A (Primers/probe) into each well of Row A, and the first two wells of Row B, and then pipet 7.12 μL of Tube C 10⁻², 10⁻³, 10⁻⁴, 10⁻⁵, 10⁻⁶, 10⁻⁷, 10⁻⁸ (curve DNA + reaction mix) into the wells as indicated (Fig. 2).

10. Pipet 16 μL of Tube B primers/probes/reaction mixture into all empty wells (Row B3–B12, Rows C-H, Fig. 2), and then pipet 4 μL diluted sample or control DNA into these wells, as indicated.

Table 2.

Preparation of Real-time MSP Reaction Mix Sufficient for 72 sample wells and U, M and Water Controls

Real-time MSP (STEP 2) Reaction Mix		1× (μL)	110× (μL)	Final concentration
Reaction mix	10× MSP buffer	2.00	220.0	1×
	25 mM dNTP (Denville Scientific)	0.16	17.6	200 μM
	ROX 50× (Life Technologies)	0.24	26.4	300 nM
	10 mg/mL tRNA (Roche)	0.10	11.0	50 μg/mL
	5 U/μL RAMP Taq (Denville Scientific)	0.20	22.0	1.0 U
	Water	0.42	46.2
	Reaction mix subtotal μL	3.12	343.2
Primes/probes	Internal M primer, 5 μM forward	3.20	352	700 nM
	Internal M primer, 5 μM reverse	3.20	352	700 nM
	Internal U or primer, 5 μM forward	3.20	352	700 nM
	Internal U or primer, 5 μM reverse	3.20	352	700 nM
	M probe, 100 μM (FAM/TAMRA)	0.04	4.4	200 nM
	U or probe, 100 μM (VIC/TAMRA)	0.04	4.4	200 nM
	Primer/probe mix subtotal μL	12.88	1416.8
DNA	DNA (diluted 1:10−2 or 1:10−4)	4.00
	Total μL	20.0

1 Full plate	Calculated amount	Reaction mix (Rxn)	Primers/Probes
Standard curve	(7 dilutions)(3× rxn) = 21×	(21×)(3.12) = 66 μL	(14×)(12.88) = 180 μL
Samples	(82×) + (3× extra) = 85×	(85×)(3.12) = 265 μL	(85×)(12.88) = 1095 μL
Total	Amount needed	106× + extra = 110×	99× + extra = 110×

Fig. 2.

The QM-MSP plate setup schema

3.9. Performing the Real-Time MSP run

1. The real time PCR reaction is run a final 20 μL volume containing 16.6 mM NH₄SO₄, 67 mM Tris pH 8.8, 6.7 mM MgCl₂, 10 mM β-mercaptoethanol, 0.1% DMSO, 300 nM ROX, 200 μM dNTP 5 μg/mL tRNA, 1.0 U RAMP Taq polymerase, 700 mM each forward and reverse primer, and 200 nM for each hydrolysis probe.

2. On ice, secure an optical adhesive cover on the 96-well PCR plate, taking special care to seal all edges well to prevent evaporation.

3. Centrifuge the plate in the cold at 250 × g for 3 min.

4. Do not contaminate the underside of the plate with any laboratory dyes such as ethidium bromide or the PCR block will become contaminated.

5. Using the software provided by the manufacturer, specify the PCR as the absolute quantitation method, label wells for standards, unknown samples, and controls, identify which probe in conjugated to which of the two fluorophores, and indicate the presence of the ROX passive dye.

6. Specify parameters as 95 °C for 10 min, then 40 cycles of 95 °C for 30 s and 65 °C for 1 min (see Note 15).

7. Specify the volume in the wells as 20 μL.

8. Transfer the plate to the real-time PCR machine and start the run. Approximate time to completion is 90 min. Before the run the plates may be stored refrigerated in the dark up to 24 h without any loss of signal.

3.10. Analysis Calculating Gene Methylation

Using the Applied Biosystem’s 7500 Real-time PCR System, sample values are extrapolated from the standard curve for target and reference DNAs. This is called absolute quantitation. It is performed according to the instructions from the manufacturer, with several modifications. Use the software supplied to analyze the data.

1. With the baseline set at automatic, manually set a single threshold for all wells at the point of maximal overlap of the curves for methylated TARGETgene (M) and REFERENCE (U gene) (usually ΔRn between 0.02 and 0.04).

2. For the standard curve, assign the Ct for the 10⁻² dilution = 200,000,000 copies. The Ct for the 10⁻³ dilution is assigned 20,000,000 copies, the 10⁻⁴ Ct is assigned 2,000,000 copies, etc. with the final 10⁻⁸ Ct assigned 200 copies.

3. If M and U do not perfectly overlap over the entire range of the standard curve, the copy number is adjusted according to the average ΔCt of the two curves.

4. Calculate the ΔCt between Ct of methylated TARGETgene and Ct of unmethylated REFERENCE gene DNAs at each dilution.

5. Calculate the overall average ΔCt between the two curves, and then transform the ΔCt value to fold-difference in copies (see example below).

6. Adjust the copy number of the highest dilution curve based on this factor. Example, if the methylated TARGETgene curve was at the lowest dilution (had the lower Ct values) and the average ΔCt between both curves was 1.5 cycles (considering all dilutions), the difference would be 2^1.5 = 2.828 fold.

7. Assign 200,000,000 copies to the 10⁻² dilution of the methylated TARGETgene and 70,700,000 copies to the 10⁻² dilution of the REFERENCE gene (200,000,000 copies ÷ 2.828 fold = 70,700,000 copies). After the standard curve Ct values are assigned copy numbers, use the software to extrapolate from the curve to the samples the copy number of M and U gene DNA.

8. Use the formulas below to calculate the extent of methylation for each gene separately, and cumulatively for the gene panel (Examples in Figs. 3 and 4): QM-MSP: Percent Methylation (%M) = [copies Methylated TARGETgene ÷ (copies Methylated TARGETgene + copies unmethylated TARGETgene) copies] [100]; CMI = the sum of all %M values within the panel.

9. If the run fails to meet minimum criteria, rerun STEP 2 after making appropriate corrections.

   1. Standard curve: R² = ≥0.99, efficiency = 90 ± 10% and slope = −3.33 ± 10%. TARGETgene and REFERENCE gene C_t overlap ≤1.5 cycles, averaged over the whole curve.
   2. For any sample, copy number of methylated TARGET gene or REFERENCE gene must not exceed the upper range of the standard curve (200,000,000 copies, 10⁻² dilution of the curve) and the copy number of the REFERENCE gene must be at least 20,000 copies.
   3. Control water (after both multiplex and real-time PCR), C_t ≥ 38.
   4. Control fully methylated TARGETgene DNA (universally methylated DNA template, after multiplex and real-time PCR), %M or MI = 100.
   5. Control REFERENCE DNA (fully unmethylated allele of TARGETgene after multiplex and real-time PCR), %M or MI = 0.
   6. Average duplicate samples for patient samples, duplicate aliquots are assayed and averaged.
   7. Implausible extreme values (e.g., due to contamination) are discarded/repeated.

Fig. 3.

QM-MSP Assay of Breast Cells collected during ductoscopy. (a) In women presenting with spontaneous nipple discharge, QM-MSP was performed for a panel of six genes on DNA from ductal cells retrieved during ductoscopy. (b) Lesions of concern observed during ductoscopy were surgically excised and classified (X-axis) according to the histological diagnosis: DCIS (ductal carcinoma in situ), INV (invasive ductal carcinoma, ADH (atypical ductal hyperplasia), UDH (usual ductal hyperplasia), or Pap (papilloma). Procedure: Ducts with minimal findings (Minimal) during ductoscopy were not subjected to surgery; in cells derived from these ducts methylation was low (90th percentile = 6.2 units of cumulative methylation). Methylation was significantly higher in cells derived from ducts containing DCIS and INV cancer compared to those cells obtained from ducts with from papilloma (benign) or ducts with minimal findings (Mann–Whitney, p = 0.002 and p = 0.0001, respectively). Each colored segment represents the % methylation of a different gene, and cumulative methylation is represented as the sum of methylation of all genes (Y-axis). (Inset) Approximately 30 cells were adequate to quantitate the level of methylation for the six-gene panel in a DCIS in (b). Figure adapted by permission from the American Association for Cancer Research: Fackler et al., Hypermethylated genes as biomarkers of cancer in women with pathologic nipple discharge. Clin Cancer Res. 2009; 15:3802–11 [10]

Fig. 4.

STEP 2 amplification curves and analytical parameters. For the RASSF1 gene, amplification plots are shown for methylated TARGET (M) and REFERENCE DNAs, as indicated by arrow. Standard curve amplification plot demonstrates linear overlapping M and REFERENCE U curves ranging from 200,000,000 to 200 copies, with amplification efficiency of 97.6% and 96.5%, respectively. Figure adapted by permission from the American Association for Cancer Research: Fackler et al., Quantitative multiplex methylation-specific PCR assay for the detection of promoter hypermethylation in multiple genes in breast cancer. Cancer Res. 2004;64:4442–52 [2]

4. Notes

• 1Stocks of 1 M Tris–HCl (pH 8.0), 5 M NaCl, 500 mM EDTA (pH 8.0), 10% SDS, TNES, 1 M MgCl₂ are stable at room temperature indefinitely. Salmon sperm and tRNA stocks are stable at −20 °C indefinitely. Proteinase K should be kept at −20 °C until the expiration date specified on the tube. 10× MSP buffer is only stable for 1–2 months at −80 °C. 1 M NH₄SO₄ should be prepared immediately before using to make 10× MSP buffer. Purchase fresh beta mercaptoethanol every 6–12 months and keep it in the −20 °C freezer.
• 2Keep the 10× MSP buffer stocks at −80 °C. This buffer goes “off” after a few months.
• 3Example: Design of AKR1B1 QM-MSP primers/probes.

  The AKR1B1 gene was first reported to be hypermethylated in breast cancers based on our previous methylome array studies [15, 16]. The methylation status was verified by QM-MSP assays using primers/probes located within 100 bp of the array probe [15, 16].

Wild type genomic DNA sequence of AKR1B1

gcagctgaggaactcctttctgccaCGCGgggCGCGggCGagCGttgggg gCGgaaagaatcCGctgccactaggaccaggCGgaagaagcatcccCGcCGacccttggggaaggcCGcCGCGgcacccccagCGcaaccaatcagaaggctccttCGCGcagCGgCGCGccaacCGcaggCGccctttctgcCGacctc aCGggctatttaaaggtaCGCGcCGCGgccaaggcCGcacCGtactgggCGggggtctggggagCGcagcagccatggcaagcCGtctcctgctcaacaaCGgCGccaagatgcccatcctggggttgggtacctggaaggtaggtgctCGtgggggCGCGggccCGgggctCGcctcacactctcCGCGCGgcctgtat tggCGagggacccCGagtgaccctgagcagctCGcccCGCGgaCGccCGg

Sequence of bisulfite-treated methylated AKR1B1 DNA

gtagttgaggaatttttttttgttaCGCGgggCGCGggCGagCGttgggggCGgaaagaattCGttgttattaggattaggCGgaagaagtattttCGtCGatttttggggaaggtCGtCGCGgtatttttagCGtaattaattagaaggttttttCGCGtagCGgCGCGttaatCGtaggCGttttttttgtCGattttaCGggttatttaaaggtaCGCGtCGCGgttaaggtCGtatCGtattgggCGggggtttggggagCGtagtagttatggtaagtCGttttttgtttaataaCGgCGttaagatgtttattttggggttgggtatttggaaggtaggtgttCGtgggggCGCGggttCGgggttCGttttatatttttCGCGCGgtttgtattggCGagggatttCGagtgattttgagtagttCGtttCGCGgaCGttCGg

Sequence of bisulfite-treated unmethylated AKR1B1 DNA

gtagttgaggaatttttttttgttaTGTGgggTGTGggTGagTGttgggggTGgaaagaattTGttgttattaggattaggTGgaagaagtattttTGtTGatttttggggaaggtTGtTGTGgtatttttagTGtaattaattagaaggttttttTGTGtagTGgTGTGttaatTGtaggTGttttttttgtTGattttaTGggttatttaaaggtaTGTGtTGTGgttaaggtTGtatTGtattgggTGggggtttggggagTGtagtagttatggtaagtTGttttttgtttaataaTGgTGttaagatgtttattttggggttgggtatttggaaggtaggtgttTGtgggggTGTGggttTGgggttTGttttatatttttTGTGTGgtttgtattggTGagggatttTGagtgattttgagtagttTGtttTGTGgaTGttTGg

AKR1B1_F_Ext	gYGtaattaattagaaggtttttt	Tm 58	216 bp
AKR1B1_R_Ext	aacacctaccttccaaatac	Tm 56
(Gtatttggaaggtaggtgtt sense)
AKR1B1_FM	gCGCGttaatCGtaggCGttt	Tm 64	84 bp
AKR1B1_RM	cccaataCGataCGaccttaac	Tm 64
(gttaaggtCGtatCGtattggg sense)
AKR1B1_FUM	TGgTGTGttaatTGtaggTGtttt	Tm 64	86 bp
AKR1B1_RUM	cccaataCAataCAaccttaacC	Tm 64
(GgttaaggtTGtatTGtattggg sense)
AKR1B1_M_Probe	CGtacctttaaataaccCGtaaaatCGa	Tm 76
(tCGattttaCGggttatttaaaggtaCG sense)
AKR1B1_U_Probe	ACAtacctttaaataaccCAtaaaatCAac	Tm 76
(gtTGattttaTGggttatttaaaggtaTGT sense)

• 4.Often, not all design criteria can be met, so factors must be balanced against the sequence in the location that is targeted. CG content and Tm are the most critical primer criteria.
• 5.TARGETgene and REFERENCEgene for one analyzed region will be amplified in the same well as a duplex reaction requiring thus a FAM/TAMRA labeled hydrolysis probe for the TARGETgene and a VIC/TAMRA labeled hydrolysis probe for the REFERENCEgene (or vice versa, see also Fig. 1).
• 6.For prolonged storage, these DNAs are routinely diluted in 1X Dilution Buffer containing a large excess of salmon sperm and tRNA in Tris buffer, pH 8.8. This stabilizes the DNA, and prevents losses due to degradation and adsorption to the plastic microcentrifuge tube.
• 7.One of the most important issues when working with small amounts of DNA sample is to add carrier such as salmon sperm DNA and/or tRNA to the lysis buffer to prevent nonspecific losses of template DNA. However, if the sample will also be used for methylation array or DNA sequencing do not use carrier.
• 8.Before scraping, prewet the tissue with a couple microliters of TNES so it stays stuck to the slide.
• 9.Samples stained with Wright’s stain may not amplify well by MSP.
• 10.For the 42 °C incubation with M-dilution buffer in the sodium bisulfite conversion protocol, prepare the samples on ice and at the end of the incubation transfer the samples back to ice. This reduces sample degradation. Do not incubate samples longer than specified by the protocol. Similar precautions need to be taken during desulfonation of samples. Take into consideration the amount of time needed to process all samples when timing incubations. Neutralize the reaction with M-wash buffer immediately after removal of Desulfonation Buffer from the column.
• 11.Do not centrifuge the columns longer than specified times or the column may become overly dry. This may reduce DNA recovery.
• 12.Small bisulfite crystals may form during the overnight sodium bisulfite conversion step, but they will not interfere with DNA binding to the column after the addition of M-binding buffer.
• 13.For clean-up of DNA after the sodium bisulfite conversion step the DNA sample must be thoroughly mixed with the M-binding buffer before centrifugation of the column, otherwise the recovery of DNA may be low due to poor loading of the column.
• 14.Vortex all reagents, primers, and probes after thawing. However, do not vortex stocks of Taq polymerase. In STEP 1, gently mix by stirring with a pipette tip between the addition of all components, followed at the end with a pulse-vortex and brief centrifugation of the master mix.
• 15.Annealing/elongation temperatures may need to be optimized (ranging from 60 to 65 °C) depending on the primers.