Over the past two months, we have been testing various reaction conditions for the ABI Big-Dye Terminators with double-stranded DNA templates ranging from pUC or pGEM without any insert up to bacterial genomes in the 2 Mbp size. Below are the results of these experiments and the conclusions we have reached. PLEASE BE ADVISED that any deviation from the ABI protocols as published in ABI's literature is not supported by ABI. I only am providing the information below so you will have an idea of what we have done but and the disclaimer is "use these protocols at your own risk". If you like them and are happy you learned about them, send me some e-mail. If they don't work for you, keep it to yourself... :-)
Also please be aware that 1 ul of the Thermofidelase enzyme does cost and therefore THERMOFIDELASE ONLY SHOULD BE USED SPARINGLY!! Thus, if you do need to use it, you can dilute the Thermofidelase enzyme with 5x reaction buffer at least 4-fold in case you need it (see note below). Finally, there is no reason to even think of using Thermofidelase routinely for any sequencing other than bacterial genomic. However, if you have the "gap from hell" or the "repeat from hell" then it may be worth a try along with increasing the extension temperature to 72degC, increasing the DMSO concentration, etc.
If you have any suggested changes to the protocols below, please let me know what experiments you have done and the results, so that I can implement them.
Thanks and happy sequencing,
++++++++++++++ Big-Dye Protocols and Notes +++++++++++++++
Below are the sequencing conditions we have tested for the ABI Big-Dye terminators (PE-ABI #4303150 for the 1000 reaction kit - Description: TF,KIT BTD RR-1000) with various templates. As usual, it always is important to quantitate all templates by agarose gel electrophoresis vs size and concentration standards and do a few tests with different template concentrations to determine the optimal conditions for your reactions.
Although several conditions are given below, the ones we routinely use are underlined with ++++++'s. Notice also that you SHOULD NOT REDUCE either the absolute amount of primer or template when the reactions are scaled down.
============= For Cosmids, BACs, PACs, Fosmids =============
|Reaction "size"||Big Dye Mix (ul)||Primer (pmoles)||Template (ngrams)||H20 to final volume of||Num. of Cycles|
|TIGR's 1.5x reaction||12 ul||10-30||400-1000||30 ul||25|
|ABI 1x reaction||8 ul||10-30||400-1000||20 ul||45|
|OU 1/2 reaction||4 ul||10-30||400-1000||10 ul||60 - 99|
|OU 1/4 reaction||2 ul||10-30||400-1000||5-7ul||99 (see Note #3)|
Note #1: The results are improved using a modification of the original bacterial thermo-cycling conditions obtained from Cheryl Heiner rather than the FS cycle sequencing conditions, and by increasing the number of cycles to 100 overnight. Typically we are using 18-22 mers as primers where no more than 65% identity (i.e. <14/20 Smith Waterman Identity) with all the known sequences in the project database, including cloning and sequencing vectors. With longer primers, e.g. 30 mers, the annealing temperature and extension temperature each should be raised 5degC to 55 and 65degC respectively to reduce non-specific binding.
Note #2: All of the above reactions work with 5 to 10% DMSO, with or without Thermofidelase. However, since 1 ul of the Thermofidelase stock is recommended for a 30 or 20 ul reaction, the Thermofidelase stock should be diluted with 5x rxn buffer such that:
1 ul of a 1:2 dilution should be used in the 1/2 reaction, and
1 ul of a 1:4 dilution should be used in the 1/4 reaction.
Note #3: The Modified Cosmid, BAC, PAC, Fosmid Thermo-cycling Conditions
Note #4: Increasing the number of cycles from 45 to 60 to 80 and eventually to 100, almost linearly improved both the signal and read length. Increasing the annealing temperature from 50 to 55degC and the extension temperature from 60 to 65degC also improved the signal and read length but to a lesser extent than increasing the number of cycles. Because of the increased number of cycles, these incubations only should be done overnight.
Note #5: ThermoFidelase improves the signal and read length with bacterial genomic templates for all primers tested.
Note #6: Although some primers work better than others and we have no clue why, it is clear that any proposed primer should be screened against the already known sequence data for the target cosmid, BAC, PAC, Fosmid clone, as well as sequencing vector sequences. Thus, to prevent the primer from binding to multiple places on the target clone, the primer sequence should be compared to all the contigs in the target clone database. Only those primers with a homology of less than 12/20 Smith Waterman Identity elsewhere on the target clone should be used for custom primer directed sequencing. (i.e. Thus, to eliminate multiple priming, the primer, which has 100% sequence identity to the site you want it to bind, should not have more than 60% homology elsewhere in the target large insert clone.)
The PrimOU computer program is available from our informatics group that has been modified from the SW Medical Center's Primo program, to allow screening against multiple contigs in a phred/phrap database at various stringencies. In addition, typically we choose primers with either a G or C on their 3'end and a roughly even distribution of each of the 4 bases or a slightly higher G/C content (i.e. 1-2 more G's or C's than A's or T's to shift the Tm slightly higher)
Note #7: As one increases the primer length, the annealing temp. can be increased so too can the extension temperature. We have tested 20,25,30,35 mers on bacterial genomic templates as well as increasing the extn temp from 60 to 65 to 72degC and the annealing temperature from 50 to 55, 60, 65, 70degC.
Our conclusion is that:
*PE-ABI #4303150 the 1000 reaction kit: Description: TF,KIT BTD RR-1000
**Fidelty Systems Inc. (301) 527-8250
Bruce Roe, firstname.lastname@example.org