Epigenetics - appunti di lezione

Dna Methylation Assessment

Techniques overview

The techniques can target: single CpG, global DNA methylation, or genome-wide methylation. DNA methylation techniques can be divided into two groups from the technical point of view:

Direct approaches

These quantify the number of methyl groups contained in a DNA sample. They are chemical techniques (Liquid Chromatography or Capillary Electrophoresis) and are used only for global methylation analysis, not very often used in biology.

Indirect approaches

These rely on specific recognition of DNA-methylation loci and rely on amplification of DNA, of a specific area of the DNA, or of all the DNA that we want to analyze. They are based on different approaches:

• Bisulfite treatment: the most spread now.
• Oldest techniques: based on restriction enzymes
• Antibodies used to recognize specifically methyl groups

How to choose the technique?

1) Focus on what you’re going to investigate: a single CpG, a global analysis, or several CpGs, one near to another.

2) Consider, for each technique, advantages and limitations.

3) It also depends on the facilities and technologies available in your lab.

How to identify CpGs?

Follow the 5’-3’ direction to check how many CpGs there are (P is the phosphate group). They are generally symmetric. C of CpG can be methylated.

Single CpG analyses

If you want to study only a CpG in a sequence, you can rely on two main approaches:

• Restriction enzymes, followed by techniques to visualize and understand the results of the first part
• Bisulfite treatment, followed by techniques to visualize and understand the results

Enzymatic restriction of genomic DNA

Restriction enzymes, which are endonucleases isolated from bacteria, derive their names (e.g., EcoR1) from the genus, species, strain, and order of identification. They can cut the double strand DNA producing blunt ends (even cut) or sticky ends (a strand is protruding), recognizing specific sequences called restriction sites, which are palindromic sequences (there’s a symmetry in the sequence, different from mirror repeat). Some of them are isoschizomers: different restriction enzymes can cut the same restriction site. They are used to study DNA methylation: Msp1/Hpa2 recognize the same palindromic sequence, CCGG, but Hpa2 can cut only if the DNA is not methylated, Msp1 can cut every time independently of the methylation state of the internal C, so it is an important internal control. They can be used to differentially digest a sequence of DNA based on its methylation because their restriction site contains the CpG, which is our target.

Considering this sequence with more than one restriction site, Hpa2 digestion will cut only unmethylated C and won’t cut methylated C, Msp1 digestion will produce 3 cuts, cleaving every CpG sequence. So, you have to perform digestion in parallel with both enzymes separately to distinguish if the cut occurred or not. Msp1 digestion is important as an internal control that confirms the presence of the sequence of interest.

Visualization of the results

• Southern blot: A huge amount of DNA is required: 10 µg of DNA are digested overnight with a first enzyme (as BamH1) that cuts the region of interest and with Hpa2, another 10 µg are digested with the first enzyme and with Msp1. Digested DNA is loaded in another gel and electrophoresis will separate the DNA based on the fragment length. Then the digested DNA is transferred to a membrane where you can visualize where your target of interest is by using a probe (the probe should not be very long to understand better if the cut occurred or not). Msp1 digested sample should represent a complete digestion of DNA (control). Determining the size of the bands in Hpa2 digested samples allows finding the restriction sites. This method, anyway, is very time and labor demanding (it’s not applied anymore currently).
• PCR: Digested DNA can be amplified by PCR.
  • Design of PCR essay: 2 primers to anneal the sequence of interest, DNA polymerase, and all the nucleotides for the reaction.
  • Load the products in agarose gel (electrophoresis), that separates DNA based on the dimension of the fragments (smaller the DNA molecule and faster the migration and longer the pathway). Primers will give PCR products only if the C is methylated and not cut by Hpa2, otherwise, they won’t be able to amplicons.

  Example: Methylated site => Hpa2 cleaved => You find amplification products by running the gel. Unmethylated site => the enzyme cut => no products (amplicon) in the gel.

  So, digestion with restriction enzymes amplification by PCR agarose gel. Controls can be added to check if a negative result is due to unmethylated sequences or failure in digestion reaction. 100% methylated DNA and 0% methylated DNA to create a calibration curve and so you can have an idea of the methylation of your specific sample, but it is not very accurate. Nowadays there are more accurate methods that can be applied.

Weak points for all the restriction enzyme-based techniques:

• Methylation can be detected only in CpGs located within the tetranucleotide recognition sequence
• Southern blot is quite complex because with many CpGs that are cut, it’s difficult to visualize in a probe in the membrane the area of interest
• Incomplete digestion might give false positives (ER are not 100% efficient, there are specific T and times, there could be random cuts or incomplete cuts).

Bisulfite treatment of genomic DNA

Sodium bisulfite is a chemical that converts only unmethylated cytosines into uracils (C→U), in 3 steps: sulphonation, hydrolytic deamination, desulphonation. Methylated cytosine won’t be converted (because mC impairs the 1st step). PCR amplification causes the conversion of U to thymine (U→T) because U is not a DNA base. Amplification must be done after the conversion, in order not to lose information about the methylation state: in in vitro amplification there is not the action of maintenance DNMT1, so during replications, the methylation pattern at hemi-methylated sites won’t be repristinated.

Detection techniques to analyze the results

• Methylation-specific PCR: It is possible to design 2 primers, one designed to anneal with the unconverted methylated C (having G that pairs with C in its sequence), the other designed to anneal with the converted unmethylated C (having A that pairs with T). These 2 primers are used for amplification in 2 different tubes, then the results are observed by electrophoretic run:
  • If only the primers designed for unconverted C give a product => C was methylated
  • If only the primers designed for converted C give a product => C was unmethylated
  • Both primers give a product: half C methylated, half not.

  Limitations: the method is qualitative* (just a quantitative estimation) and there is risk for false positives and false negatives, even if the use of controls improves the situation.

• Real-time PCR MSP: It is based on the use of 2 different primers, one specific for unconverted mC, one for the converted C. By comparing the different cycle threshold (cycle at which the signal is detected) it is possible to evaluate the amount of the starting DNA (the lower the ct, the higher the starting amount). It allows a better quantification of methylated and unmethylated samples. Unspecific products that can form during the amplification, as primer dimers, can be detected using the melting curve. This is generated by exposing DNA at increasing T until it is denatured. Because the T at which DNA is denatured is sequence-specific (A-T requires less energy than C-G to be broken), peaks not corresponding to the amplicon of interest are not considered. Advantage/disadvantage: it’s a quantitative* analysis with possible unspecific peaks that anyway can be detected through melting curves.
• Methylight: It is a real-time PCR based on the use of Taqman probes: the essay requires 2 primers (a forward and a reverse) and 2 probes that have a fluorophore at the 5’ and a quencher at the 3’. Two fluorophores are associated with 2 different probes (consisting of a ssDNA), one able to bind the sequence with converted C, the other able to bind the sequence with unconverted mC. As the quencher flanks the fluorophore, it inhibits its signal, but when the polymerase elongates the primer, it is able to cleave the fluorophore thanks to its 5’-3’ exonuclease activity: the fluorophore emits the signal. Depending on the kind of signal that is detected, we are able to establish if DNA was methylated or not: the fluorescence from the released fluorophore is proportional to the amount of accumulated PCR product. The quantification of the degree of methylation will be based on the threshold cycles registered for the fluorescence of the two probes. Advantages: This technique is better than classic PCR because is more specific, with reduced risk of false positives, different loci could be tested simultaneously using different labeled probes.
• COBRA (combined bisulfite restriction analysis): It combines bisulfite conversion - PCR and restriction enzymes. DNA conversion with bisulfite PCR amplification digestion of PCR products with restriction enzymes electrophoresis. The restriction enzyme is able to cleave originally methylated sites (CGCG), leaving sites originally unmethylated (TGTG). Unmethylated locus => no cleavage, methylated locus => cleavage. If the conversion changed the sequence, the cut will not occur later (this is the main point). Using the cleavage pattern of the PCR products, the methylation state of the original locus sequence can be determined.
• Methylation sensitive high-resolution melt analysis: It uses a melting curve obtained by exposing DNA to increasing T (usually by heating), after the bisulfite conversion. If the conversion occurred, the unmethylated C was converted to T: A-T bond is easier to be broken with respect to C-G bond obtained in case of methylated unconverted C. The difference in sequence even of a single nucleotide determines a different melting T necessary to denature the DNA molecule. This is exploited by the essay, using saturating intercalating dyes as Eva green (the most used, sometimes also SYBR green) that are able to give the signal until the DNA is double-stranded. Once the denaturation occurs, the DNA is single-stranded, and the signal is not given anymore. Results: In case of a clear pattern (DNA fully methylated or fully unmethylated) the results are very clear, in case of a mixed pattern of methylation, they are difficult to interpret. This technique requires always controls, using 100% methylated and 0% methylated controls to contextualize the curves in the middle. This technique can also be used to detect mutations in DNA.
• Pyrosequencing: It represents the gold standard: the best technique to perform single CpG analysis. It’s a sequencing by synthesis technique. The DNA sequence is determined by light emission upon incorporation of the right complementary nucleotide: only one of the four possible nucleotides is added at a time, until the right one is incorporated and emits light. In the presence of APS (adenosine 5’ phospho-sulfate) ATP sulfurylase converts pyrophosphate (released from the incorporated correct nucleotide) to ATP, which is used by luciferase to convert luciferin to oxyluciferin generating visible light. The light is detected by sensors and transformed into a peak in the pyrogram. Each peak corresponds to a nucleotide and the height is proportional to the number of nucleotides added in that position. Apyrase continuously removes unincorporated nucleotides and ATP. When degradation is complete, another nucleotide is added. Bisulfite conversion PCR amplification: using primers that don’t overlap the CpG of interest and one of the 2 primers must be biotinylated to isolate only one single strand of the amplicon that is used for the sequencing. This strand is isolated by the workstation, able to denature DNA of PCR products, and only the biotinylated strand is retained by the beads. Then the sequencing occurs by synthesis using sequencing primers for the synthesis of the complementary strand. Result interpretation: By reading the sequence, if you find a T it means that C was not methylated, if you find a C it’s because it was methylated. Usually, there’s not fully methylated or unmethylated DNA and with this technique, you can fully quantify the amount of methylated or unmethylated amplicons present in your reaction. It will put one nucleotide each time: before the T and try to amplify, quantify, and identify how many amplicons will anneal with the T, then remove the T, put the C in the reaction and then identify the number of amplicons that include the C and so you can have the exact quantification of the methylated and unmethylated of each CpGs. Advantages:
  • The method is much more specific.
  • It gives quantitative info on the % of methylation
  • It can analyze all the CpGs, even many CpGs one near to the other, differently from ER.
  • It is very accurate with a sensitivity of 2-5% which can be increased by making duplicates.

Questions

• How can you use restriction enzymes to analyze DNA methylation? Which enzymes are mostly used and why? (Enzymatic method and COBRA)
• In the PCR performed after the enzymatic digestion, are the primers methylation-specific or not? It’s not necessary that they are sensitive, because they flank the region of interest which is cut or not.
• In DNA methylation analysis, what is bisulfite used for? Only if the C is not methylated is converted.
• In the PCR after the bisulfite conversion are the primers methylation-specific or not? Yes, they must be.
• Why can melting temperature measurement be used to study DNA methylation?

Global Genome Methylation Essay

It consists of the measurement of the average methylation level of the genome, without defining where it occurs (that is the goal of the whole genome methylation analysis). It relies on 2 methods:

ELISA essay

It uses antibodies and color change to detect the presence of a certain antigen, in this case, the antibody recognizes the methyl group, allowing the quantification of the total amount of methyl groups. There are 2 cases:

• The antibody is fixed on a well surface
• The DNA is fixed on the well surface (most frequent).

After denaturation, DNA is incubated, fixed, and blocked in the surface well. Then, incubation with primary antibody (anti-5-methylcytosine): it recognizes the methyl group, and its amount depends on the methylation. To visualize the primary antibody, we perform the incubation with a secondary antibody conjugated with an enzyme that gives a light signal that is read by a spectrophotometer at a certain length (usually 420 nm) to determine the absorbance intensity, dependent on the methylated C. In fact, the light depends on the amount of secondary antibody which is dependent on the amount of primary antibody which is bound to the methylated C, and we can quantify the amount of methylated C in the genome comparing different samples.

Pyrosequencing LUMA (Luminometric Methylation Essay)

This technique includes restriction enzymes and pyrosequencing to count the cuts that are generated by them. Three restriction enzymes are used:

• EcoR1: a sort of normalizer, its restriction site is GAATTC leaving 5’AATT protruding complementary ends
• Hpa2 and Msp1: used to differentially digest methylated and unmethylated C, they leave 5’-CG protruding ends after the cleavage.

Two parallel digestions are performed for each sample: the first with Msp1, the second with Hpa2; EcoR1 is added in both. After the digestion, we perform pyrosequencing, that won’t give the sequence of DNA, but duplicates the end that remains protruding after the cut. Considering EcoR1 cut, only A-T are added at its protruding end. Then C and G are added to complete the polymerization of the ends created by Msp1 and Hpa2. Pyrosequencing allows seeing how many A and T are necessary to duplicate all the cut performed by EcoR1 and how many C and G to complete the cut performed by Hpa2 and Msp1. It gives 4 peaks: A, T, C+G, and C+G that is control.

EcoR1 represents a control for the amount of DNA that must be the same in the 2 wells, treated with different enzymes: A (yellow) and T (green) must have, respectively, the same height in both digestions, otherwise, it means that more DNA was put in one of the 2 wells. Considering the CG peaks (blue), they are different in Hpa1 and Msp1 digestions. Considering the C+G peak of Hpa2 digestion, the enzyme couldn’t cut the same restriction sites cut by Msp1: this means that some cuts didn’t occur because CpG were protected from the methylation. If instead, you would have a peak as high as the one obtained by Msp1 digestion, it meant that DNA was completely demethylated.

The percentage of methylation can be calculated following the formula: DNA methylation is defined as the Hpa2/Msp1 ratio: the higher the ratio, the lower DNA methylation, and the contrary. If DNA is completely methylated the ratio would be 0, if it is completely unmethylated it is 1. The incorporation of biotinylated dCTP is thus directly correlated with the number of digested sites and inversely correlated with DNA methylation.

LUMA pros and limitations

• Pros:
  • LUMA provides better control of results than ELISA, that lacks the control for the amount of DNA in each spot, even if you put controls of completely methylated or unmethylated, sometimes the results are not sensitive.
  • LUMA provides an absolute quantification without needing completely methylated and demethylated DNA controls.
• Limitations:
  • Quite a high amount of DNA needed and use the restriction enzyme (if they don’t cut properly, you can have some risk of false positives).

Genome-wide methylation

It is based on the analysis of genome-wide methylation, considering every CpG of the entire genome. The analysis is organized into two steps that can be performed by different methods: genomic fractionation and analysis of methylation patterns that can be performed by several methods:

• Genomic fractionation through different methods
• Analysis of methylation patterns to read the results, by microarray or NGS.

It is important to choose the proper techniques and consider the facilities.

Restriction endonuclease digestion followed by microarray analysis

Details on this method are not provided in the original text.