Increasing Annealing temperatures for BSP-PCR - New Paper in Biotechniques (Feb/16/2007 )
Hi all,
anyone seen this paper (Chen et al. Optimizing annealing temperature overcomes bias in bisulfite PCR methylation analysis, Pubmed-ID: 17269485)?
The main statement of the paper is, that the efficacy of PCR for bisulfite treated samples is better, when you use high amplificatio temperature - they used several primers with Tm around 55°C and had the best results (analyzed with pyrosequencing) at 60°C as annealing temperature!
Any ideas about that?
If one uses annealing at 60 and extension at 65°C - could it be possible to do the PCR with a two-step protocol?
Or is this a problem important only for Pyrosequencing or MALDI-TOF, as the sensitivity of "normal" sequencing is too low to be affected by this bias?
Krümel
Hi Krümel,
sorry, my institute does not have a subscription to Biotechniques 
But I'll do some guessing anyways 
In the old lab where I did the bsPCR we always tested a new primer in a temperature gradient of
(Tm - 5)°C - (Tm + 5 )°C in 1°C steps. Some primers showed more product with an annealing temp. slightly under Tm, some with
an annealing temp. slightly over Tm.
So the whole thing sounds kind of weird to me, and leads me to think it depends on the primer (what length, which bases).
Or do they mean in the paper that you obtain less product, but of a higher quality / specificity?
In this case the paper might make sense... darn I wish I could access it !
Oh - and we used the annealing temp with the highest efficacy (most product) for Pyrosequencing, and always got good results
Hi Goofy,
Biotechniques is free, you only have to register! And, your second guess was right - they had more reliable Pyrosequencing results without biasing towards unmethylated DNA...
K.
Whoops - now that's embarrasing, thanks for letting me know 
I'll take a look at the paper then 
I saw this too. the geuss is that the more methylated the DNA the more C's left in the treated DNA leading to more secondary structure. The increase in secondary structure leads to a PCR bias towards unmethylated DNA. At higher PCR temps, the secondary structure is eliminated, leading to more efficient methylation detection. I am looking into biases in methylation detection now, becuase of wierd results in the lab, look at my post please! Anyway, one thing that I should know of next week is if the use of single stranded DNA binding protien within my PCR eliminates bias for pyrosequencing (it does help with the PCR). Also looking at the amount of template used for the inital PCR. I will know more later next week.
1.) no further ideas apart from those in the paper on why exactly the bias is reducer by higher annealing temps, sorry.. secondary structure does sound good though
2.) What do you mean by two-step: annealing and extension in one step? Wouldn't this increase the bias/add a new bias?
3.) Judging by the numbers given, the bias in some of the investigated genes is easily high enough to affect normal sampling as well.
Yes, think so, too
I am not su sure. There are some reports, that PCR is more efficient in AT rich templates, when extension temp is low (65°C) - so maybe, it is worth a try? May be one next project after the multiplexing...
Sorry, I don't get your point ... what do you mean, exactly?
in 3.) I meant that the methylation of some genes investigated in the paper varies between ca. 40 % and ca. 80% (FANK1-top, FEZ1-top, P2RX5-top).
This would surely be a bias that also influences "normal" sequencing, and not only pyro-sequencing or MALDI-TOF, or not?
You're absolutely right. If this were true, it would be necessary to make a normalization with mixed templates for every tissue and every primer you use... 
I really don't like that thought 
There are other reports showing that one could overcome this bias by including one CpG into the primer sequence, near 5' end, without degeneracy! But I am not so sure about that - it's a little like trying to heat the room by setting the carpets on fire...