This phenomenon is bothering me since a while and I asked some expert advises, but did not get any satisfactory answer.
I worked with cell and protein biology in the past, but not in DNA and molecular genetics. Recently, as a part of my PhD, I am working with general and real-time PCR for expression analysis and Pyrosequencing for epigenetics (Promoter CpG methylation) studies.
For Pyrosequencing, I convert the DNA (nonmethylated C to U; deamination) with sodium bisulphite, then do PCR for the converted DNA and at last gel run to see my band of interest.
After many optimizations and standardization, I am finally being able to get accurate bands with negligible primer dimer (Very less primer dimer is acceptable). However, one interesting phenomenon I noticed was that the intensities of bands are variable?
For example, with the same sample, equal number of PCR cycles, exactly similar PCR conditions (Except annealing temperatures), equal primer concentration and with equal amount of input template DNA, the band of OCT4 promoter is less intense than that of SOX2. Likewise, the bands of NANOG and C-Kit promoters are also more or less variable in all the run. If we have 2 copies of a gene (maternal and paternal), given all conditions of PCR constant, I should get all the bands equally intense. I thought it is because of pipetting or other kind of human error, but the phenomenon repeat all the time I do the experiment. Is it because the DNA is Bisulfite converted?
(Please note that I have never compared bands for different genes for the same samples with similar PCR condition for a normal genomic DNA, as I am recently involved in molecular genetics techniques. So I have no idea if this happens to genomic DNA without any treatment?)
Thanks for your answers in advance.
Hi Prabin,
in addition to the pipetting errors you mentioned, I can see two more reasons for your results. the first is the presence of SNP at the position of the primers on one copy. the second is more difficult to review is the efficiencies between both PCR that could be be not equal. In PCR you can calculate the power of the taq to multiplicity the target by 2 at each cycle if the efficiency is at 1. but this efficiency depends on the sequence of the primers, the sequence of the target, its length, DNA state and quality (proteins??), accessibility of all stuffs ans enzymes to the sequence... if you have an efficiency of 1 for PCRA and 0.99 at PCRB, I know it's not very separate, but after 30-40 cycles you can be sure that you'll see the difference. use the DDct method for RTQ-PCR to calculate these efficiencies and incorporate them in the analysis to erase this bias.
hope this help
fred
Hi Prabin,
in addition to the pipetting errors you mentioned, I can see two more reasons for your results. the first is the presence of SNP at the position of the primers on one copy. the second is more difficult to review is the efficiencies between both PCR that could be be not equal. In PCR you can calculate the power of the taq to multiplicity the target by 2 at each cycle if the efficiency is at 1. but this efficiency depends on the sequence of the primers, the sequence of the target, its length, DNA state and quality (proteins??), accessibility of all stuffs ans enzymes to the sequence... if you have an efficiency of 1 for PCRA and 0.99 at PCRB, I know it's not very separate, but after 30-40 cycles you can be sure that you'll see the difference. use the DDct method for RTQ-PCR to calculate these efficiencies and incorporate them in the analysis to erase this bias.
hope this help
fred
I'm completely agree with Frederic, and I have to say this, intensities is not so considerable in the molecular researches for all these reasons mentioned by Frederic! Sequencing the targeted loci my be more accurate!
Except the pipetting errors, the DNA solution/might not be 'homogenous'. Any 1 uL (for example) you pipet out, might not contain exact the same DNA amount. And PCR is a very senstive tool. A little change of DNA amount can change the PCR results, and then the band intensity.
in addition to the pipetting errors you mentioned, I can see two more reasons for your results. the first is the presence of SNP at the position of the primers on one copy. the second is more difficult to review is the efficiencies between both PCR that could be be not equal. In PCR you can calculate the power of the taq to multiplicity the target by 2 at each cycle if the efficiency is at 1. but this efficiency depends on the sequence of the primers, the sequence of the target, its length, DNA state and quality of proteins, accessibility of all stuffs ans enzymes to the sequence... if you have an efficiency of 1 for PCRA and 0.99 at PCRB, I know it's not very separate, but after 30-40 cycles you can be sure that you'll see the difference. use the DDct method for RTQ-PCR to calculate these efficiencies and incorporate them in the analysis to erase this bias.
regards
Hassan,
playing the cut and paste without changing any words is not the way to do.
fred
possibilities:
1. Primers for different genes will likely have different PCR efficiencies (most likely explanation).
2. BS DNA is somewhat unstable and this may affect results especially if one gene is more affected than another or assays were performed at different times.
3. Gene Imprinting or monoallelic expression (typically only a few hundred genes are affected in this manner and therefore this biological explanation is unlikely).
4. BS converted DNA is less complex and hence it is more likely that primers are not specific for the designed region (sequence amplicons).
5. If sample is from females and genes lie on X then X inactivation may (??) be an issue (unlikely as I think Xi is largely mediated by H3K27me3 rather than methylation).
6. Always possible that genetic variation is changing primer efficiencies.
When you hear hoof beats, think horses, not zebras i.e. my bias is that the most likely explanation is #1.
Thank you everyone for your responses. I know most of the suggestions here are correct and are likely to be causing the outcome.
However, in my wildest guess (just a guess because I haven't find any reliable source of paper to support it), the promoter of the genes which are showing thin bands are more methylated. Because they are methylated, may be the methylated cytosine inhibit the activity of taq polymerase resulting in less/inhibited replication. On contrary, the thick bands are because the promoters have less/no methylation at CpG site.
This might be true because it is well established that CpG methylation at promoter reduces expression (transcription) of a gene because methylated cytosine at promoter blocks the activity of the transcription factors and RNA polymerase. If it can block the activity of RNA polymerase and transcription factors, it may also block the activity of DNA pol (Taq pol), hindering the replication process.
Please note that I have used the promoter regions of the genes, which are rich in CpG and have the ability to block transcription. Unlike the inner coding sequence of the gene, which contains negligible CpG islands, and even if they contain, researches says that they don't hamper any activity normal gene functions.
Once I finish my pyrosequencing, I will get a clear answer whether it is caused any base modification or just because of efficiency of primers and other known causes.
- Badges
- Science method