Building genes with oligos

[Steven Striepeke]

You mentioned that the majority of the "mutations" at the junctions of 
the oligos. If you were not purifying by gel I would say it was 
possibly due to significant contamination with failure sequences, this 
is particularly a problem when the capping reaction is not working well 
on the synthesizer, this leads to a lot of n-1 seq.s with almost random 
substitutions on the 5' end. Try to determine whether the errors seem 
to be toward the 5' or 3'.

As for the UV shadowing of course keep the exposure to minimum, damage to 
pyrimidines proceedes very quickly, Ts faster than Cs. Trace  out the 
outline of the   
band quickly and turn out the UV, cut the bands out under visible light.

I read something a few years ago ( cant find it now ) that suggested that 
hosts may change "unnatural" sequences. 

I used to do applications work for a DNA synthesizer co. I used to hear 
about a 
lot of replacement of C with T. After doing considerable literature 
research I theorized that Cytidine could be deaminated becoming 
deoxyuridine. When constructed into a synthetic gene fragment a G - U 
mismatch would be created. What would repair mechanisms make of that? 
Perhaps half would repair to the original seq. and in the other half the 
G would be replaced with A. I never got the opportunity to actually carry 
out experiments but what I wanted to do was use a uridine deglycosylase 
to remove the base, once the oligo contained an abasic site it could be 
cleaved with acid and be removed from possible inclusion in the construct.

Are you using any unusual protecting groups on the monomers, like FOD 
monomers, expedites, etc?

Are there a lot of Gs in the seq? They cause a lot of secondary structure 
problems?

I doubt anyone still does this but are you using DMAP in you capping 
reaction ? It can modify some Gs making them into diaminopurines.

Hopefully your post will open up a lot of discussion about the problem of 
constructing synthetic genes. Its been obvious to me from my 
conversations with synthesizer customers that it has never been as easy 
as it should be and I think the synthesizer and reagent companies would 
like the topic to stay in the closet. Not that I think they have any idea 
what the problems are but they are just afraid to have people think that 
synthetic DNA might not be perfect.

Also, do you synthesize these yourself or do you by them from a service lab?

Good Luck,
Steve Striepeke
sstriepk at crl.com

 glenn soltes 
(gsoltes at gpu.utcc.utoronto.ca) wrote: : Glenn Soltes : Cangene Corp


: Our lab has had difficulties building genes from
: complementary 40-60 nt oligonucleotides generated by the
: Applied Biosystems 392 DNA/RNA synthesizer.  A typical
: gene construction involves annealing 6-10 oligos,
: ligation, cloning into a sequencing vector, and then
: sequencing the products.  The problem is that the
: resultant genes often carry one or more mutations.  These
: mutations appear to be randomly distributed and are not
: concentrated around the junctions between the oligos. 
: The source of these mutations is not operator error or
: errors in designing the oligos because in a bank of
: clones coding for gene X, a given mutation will only
: occur once while the rest of the clones are correct.
: Transition, transversion and deletion mutations have been
: observed.

: Has anyone had any similar experiences?

: Does anyone know what is causing the mutations?
: Is it:
:  a) Probabilistic errors in the oligo synthesis.
:  b) UV exposure mutations (during UV shadowing).
:  c) Secondary structure of the oligos.
:  d) Recombination/repair

: Thank you in advance
: Glenn