Hi Michal,

In my opinion, the EpiTect Methyl II PCR Array would be a reasonable way to go in order to study a fair number of genes at one time, since you can either use a pre-made selection of "cancer" genes or design a custom array based on genes that may be of interest to you (genes involved with hematopoetic lineage decisions).

With respect to more genome-wide based approaches, the field is definitely trending toward high-throughput sequencing based approaches (RRBS, MBD-Seq, MeDIP-Seq), though array-based methods could still be considered for your project. The Illumina 450K methylation platform is a cost effective way to look at quite a number of samples (~300-400 (US$)/sample). Depending on your budget and the number of samples you might want to try, there are companies that provide next-gen sequencing based approaches do the technique and all of the initial bioinformatics for you if you would like. I have had very good luck using Zymo Research for RRBS (http://www.zymoresearch.com/services/epiquest). They provide a wide range of services from just providing the raw data back to you (if you have bioinformatic skills or access to such) or a complete bioinformatic analysis package with a list of genes that are significantly methylated or not compared to a reference (this is what I have used since I have limited bioinformatic skills myself). I know other companies are providing similar services for MBD-seq or MeDIP-seq, but I have not used them myself. Good luck with your experiments.

Hello Michal, it depends on whether you know which genes' and even cytosines' methylation states you want to assess or whether you want to discover genes whose methylation status is differentially affected in the malignant cells. In the former case PCR on bisulfite treated genomic DNA using primers that will recognize only the methylated (not converted) cytosines is an affordable method (see http://www.cellbiolint.org/cbi/imps/pdf/CBI20110649.pdf for a recent example of the application of this technique). In the second case of figure, you'd need to go for a genome-wide approach, which is technologically much more involved as you need a next generation DNA sequencing machine which is quite expensive to buy and run, not mentioning the bioinformatic expertise required to process and interpret the sequence results..

Colin

Hi Colin. Thank you for your answer and the attached paper. Would you use an in-between approach such as the EpiTect Methyl II PCR Array System http://www.sabiosciences.com/dna_methylation.php, if you want to study a limited number of specific genes?

Hi Michal,

In my opinion, the EpiTect Methyl II PCR Array would be a reasonable way to go in order to study a fair number of genes at one time, since you can either use a pre-made selection of "cancer" genes or design a custom array based on genes that may be of interest to you (genes involved with hematopoetic lineage decisions).

With respect to more genome-wide based approaches, the field is definitely trending toward high-throughput sequencing based approaches (RRBS, MBD-Seq, MeDIP-Seq), though array-based methods could still be considered for your project. The Illumina 450K methylation platform is a cost effective way to look at quite a number of samples (~300-400 (US$)/sample). Depending on your budget and the number of samples you might want to try, there are companies that provide next-gen sequencing based approaches do the technique and all of the initial bioinformatics for you if you would like. I have had very good luck using Zymo Research for RRBS (http://www.zymoresearch.com/services/epiquest). They provide a wide range of services from just providing the raw data back to you (if you have bioinformatic skills or access to such) or a complete bioinformatic analysis package with a list of genes that are significantly methylated or not compared to a reference (this is what I have used since I have limited bioinformatic skills myself). I know other companies are providing similar services for MBD-seq or MeDIP-seq, but I have not used them myself. Good luck with your experiments.

Thanks Scot, this information is very helpful.

The field word genome-wide is an approximate measure. Understanding of genome-wide methylation may be well doccumented by analyses at base pair resolution exploiting combinatination of technics, including next-gen seqing, bisulphite seqing, microarray, silenced gene specific and chromatin IP based techniques. Now, planing and use of techniques depends on the objectives. For example, in you wnat to see the expression profile of cell cycle regulatory genes, you select your genes of interest and may perform MS-PCR, whic is very common and easy today!

Michael, I can approach suggestions at depth depending on your wavelength you want!! All the best

I think it will depend on what you are going to do with the data. There is a great chance that it has already been done and you may find it already on a public database. There are very large international consortia doing just that for all type of malignancies and at different stages of cancer development. Maybe Colin knows a bit more about it. The problem with most, if not all these studies, is that we still don’t know why and how DNA methylation patterns are altered.

If you are going to look on a limited number of genes you can follow what it has been suggested above. I will add also that you should include some repeat DNA sequences, such as LINE-1, Sat2, NBL2, etc…to make it more genome wide. There are several methods to address DNA methylation on repeat sequences.

Thank you Samir , Abdelhalim and others. We want to see if the methylation status of certain genes which are known to be of importance in MDS/AML can be altered by a novel treatment approach. It is of course better if we can also assess other genes which may not yet be recognized as important in this disease. I do not have any "hands on" experience with those methods, and therefore your practical advise is very useful.

Hi Michal! If you are going to look for a limited number of genes, I think the straightforward way is to perform bisulfite sequencing. First make sure your promoters or the promoter window have CpG islands according to the accepted definition. I am sure you will already find a primers list for them in relevant publications or you can request them from the authors. Otherwise you can design them yourself using any online software. There are plenty. You can buy a kit from a number of companies such as the Epitect fast bisulfite conversion from Qiagen to treat your DNA samples, and then perform PCR, clone in TOPO-TA plasmid from Invitrogen and sequence using the provided primers from the plasmid. This is very tedious as you normally need around 20 clones per sample/ per fragment sequenced to make a statement. You will need to calculate each time the efficiency of the conversion from non CpG dinucleotides.

There are other methods but they don’t give you single base resolution or they only explore some CpG sites.

If you need to look for new genes you will need a microarray platform or next generation sequencing facility. Otherwise, you can use a candidate approach: guess which one might be.

You can also see if you treatment alter global DNA methylation which will have a great impact too. Remember, all malignancies are characterized by global DNA hypomethylation despite up-regulation of the DNA methylation machinery members. In this case you will need to quantify global DNA methylation by chromatographic means (HPLC, MS, HPCE, etc…). The choice of the method always depends on which facility you have access to. There is also a Kit from Sigma or other companies which quantify global DNA methylation by ELISA, if I remember well but in all these cases the changes could be very small which will lead to false negative results. So, t is never an easy question. You will need then to go genome wide and study also the methylation status of repeat DNA where the bulk of CpG methylation resides. I can forward you to references where to find the sequences of the primers used to analyze the methylation status of some repeat DNA and that we have used in the past.

Hi Michal, In my opinion given the significant impact of next-generation sequencing (MeDIPseq or whole genome bisulphite sequencing) in assessing epigenetic marks in cancer genomes, I would definitely advise you to give a serious thought on using this platform to look at global methylation levels. MeDIP-chip is yet another cost-effective method that you could possibly consider, coupled with bioinformatic approaches as suggested by Stephan Beck (approach termed as BATMAN. Another commonly used method is the use of mass spectrometry to look at specific genes (EpiTYPER analysis). The results generated from this platform are not only robust but moreover give you an absolute measurement of the methylation levels of specific loci. As an approach for validation of either microarray or next-gen sequencing data, I would not recommend using MS-PCR given the fact that the method utilizes information based on a short stretch of nucleotides, and in this regard it would be more appropriate to use gold standard method of bisulphite sequencing.

HI Michal,

SAB arrays or any microarray chip (Illumina, Agilent, or Affymetrix) will definitely do if you are working on regular methylation studies. Choice of NGS method should be entirely yours, in my opinion, depending upon its feasibility in your lab, its data interpretation, budget & finally conclusion which helps you to fetch some valuable result.

For your type of work any array will not be recommended as you will be able to conclude only from know sites/genes. Since you are specifically focusing on disease so work with ANY NGS based approach it will help you to screen novel methylation sites in your genome , which will be novel biomarkers, same can be used novel sites for diagnostics or treatment.

All the best