PCR with paraffin-embedded tissue DNA samples - I have had problems to amplify DNA samples from paraffin-embedded tiss (May/27/2008 )

I work with DNA samples from paraffin-embedded tissue (PET). I receive my samples from other lab, so I do not have idea how they are extracted and there is nothing that I could do about it. The thing is that I have to amplify this samples with normal PCR. All my fragments are between 150 and 300bp. I cannot split them in more pieces because I have already 45 fragments and limited amount of DNA. I have 50ul of a very bad concentration and quality DNA, and I usually dilute it in 1:10. I use HotStarTaq from Qiagen. I need this DNA for sequencing, and as it is somatic DNA, it has to be very very clean otherwise I cannot odentify my mutations. What I have tried:
1- 35-45 cycles (usually I do 40 cycles)
2- different concentrations of MgC2 (1.5 to 5.5mM) - depends of the fragments
3- DMSO and Q solution (~DMSO that comes with Hotstartaq) - good for GC rich regions
4- Different sets of primers and different primer concentration - good
5- Make a PCR from a PCR - bad
6-Different annealing temperatures.

I always do optimizations with normal DNA and find the best conditions prior to do my samples. Usually the best conditions to normal DNA is not the best for PET... What else I could do? My supervisor recently suggested do the PCR (put everything less the Taq and dNTP) and leave at 96 degrees for one hour and then do the PCR. Maybe increase annealing and extensions times too. What do you think? Do you have any suggestions?

My second problem is that recently I read a paper aying that PET DNA is error-prone and not trustable. They say they cannot confirm mutations and I have had the same problem here. Most of my mutations, even when the sequencing is very clean, could not be confirmed. What do you think about it?

Many thanks to all!
I will appreciate all efforts and suggestions.

FranPCR

The problem with paraffin embedded tissue is that the fixatives and heat used in the mounting fragment the DNA into small bits. It is also hard to get rid of the wax during extraction which affects down-stream applications.

Unfortunately there isn't much that can be done about this, other than suggest that the lab you get your samples from switches to frozen tissue and cryosections.

I suppose you could try cleaning up the DNA using a carrier such as linear polyacrylamide or glycogen to prevent further loss of DNA.