DNAseq: DyeTerminator removal for Capillary Sequencer

Bernd Weisshaar weisshaa at mpiz-koeln.mpg.de
Mon Jul 9 02:13:12 EST 2001

Dear DNA-Sequers, dear Peter,

below, there is our modification of the millipore purification 
procedure which works well in our hand (it is really close to the 
original millipore protocol). We are just moving to BDv3.0 for the 
3100, and so far the protocol seems to work as good as with v2.0. We 
may be able to use a little less kit per reaction, but we have not 
yet tested that in detail.

For a 10 ul cycle-sequencing reaction with 2 or 3 ul BD2 (depending 
on the size of the DNA template):

1: Add 300 ul water
2: Let the sephadex swell for round about 3h.
(we use G50 from Pharmacia which is loaded dry into the millipore 
setup so that the "colums" are filled completely)
3: Centrifuge 5 min at 900 g
4: Add 150 ul water (using a Hydra)
5: Centrifuge 5 min at 900 g
6: Add 9 ul water to the samples, load them **to the centers** of the 
columns and centrifuge 5 min at 900 g directly into the 96-er 
MircoAmp Optical Plates which contain 15 ul water/well (we use only 
HPLC grade water from Merck).

We only very occasionally see dye blobs, and I think that in this 
case the problem almost always is that the loading of the diluted 
sample was not done carefully to the center of the sephadex, or 
column loading with sephadex was not OK. The dilution helped in our 
hands to get constant volumes out of the plates. For the 3700, we 
"tape" the MicroAmp plates (but no v3.0 experience yet on the 3700 - 
the service just upgraded the instrument).

Best regards, Bernd

Hi Jarmo,

We use the Milliporesystem with good results. We do the following.

1: Add 300 ul water
2: Let the sephadex swell for 60 minutes.
3: Centrifuge 3 min at 1900Xg
4: Add 200 ul water
5: Centrifuge 3 min at 1900xg
6: Add sample and centrifuge 3 min at 1900xg .
7: The samples are loaded directly.

We dilute the bigdye terminitors (v2) 4 times and we do 50 pcr cycles.
In general we see only small amounts of dye blobs. As other have metioned it
is important to load the sample in the middle of the wells.
>From may to june 2001 we produced 14232 sequences (54 different users or

Average phred20: 490
Highest phred20 score: 788
Lowest phred20 score: 0
Average signal: 267
Highest signal: 3023
lowest signal 2.75

We use a 3700 so I don't know if the signals can be compared.
Has anyone tried to use this system (or a similar gelfiltration system) with
version 3 of the Bigdyes? Applied Biosystems have told me that it doesn't

Best regards,

Peter Bjarke Olsen


More information about the Autoseq mailing list