23 replies to this topic

[HomeBrew]

gyma, on Apr 2 2010, 04:39 AM, said:

if the 2nd band is specific, why does it contain human genome sequence?

HomeBrew, on Apr 2 2010, 12:10 AM, said:

What we really need is an answer to the question I posed way back here -- did you just sequence one or a few clones from a single transformation?

[gyma]

HomeBrew, on Apr 2 2010, 07:12 PM, said:

gyma, on Apr 2 2010, 04:39 AM, said:

if the 2nd band is specific, why does it contain human genome sequence?

HomeBrew, on Apr 2 2010, 12:10 AM, said:

What we really need is an answer to the question I posed way back here -- did you just sequence one or a few clones from a single transformation?

I sequenced 20 clones and they all have the same sequence, which proved to be a 2kb part of chromosome 13q.

[HomeBrew]

From a single experiment?  Could all the sequenced clones have been siblings?  Have you repeated the experiment and also recovered the same 2kb part of chromosome 13q?

Here's the point -- if your 2kb part of chromosome 13q inserts were amplicons, then they would have the primer sequences in them.  Since they do not, and it looks like the banding pattern you're seeing is a result of amplification based on your latest photo, we have to assume that the band your seeing is a mixture of amplicons (in vast excess) and co-migrating genomic fragments of the same size.  During this experimet, just due to chance, you caught a genomic fragment when cloning, and what you sequenced were all siblings.

Strange things happen when cloning -- I once recovered a segment 100% identical to a tomato gene while cloning a band from E. coli -- no idea how it happened, but I repeated the experiment, and got the segment I wanted.

There are two major issues here -- your primers are not specific enough for what you are trying to do, as they produce way too many bands, and you're relying on blunt-end cloning.  Primer specificity could be improved by re-design or changing the PCR conditions and/or cycling parameters.  Blunt-end cloning could be avoided by attaching restriction enzyme sites to the ends of your primers, or, in this case, perhaps a better solution would be to try TA cloning.  Use a Hi-Fi Taq-based enzyme, and clone into a TA TOPO vector.  Either of these methods will insure that you can only successfully clone PCR products, and take the genomic fragments out of the equation.

[gyma]

HomeBrew, on Apr 3 2010, 01:23 AM, said:

From a single experiment?  Could all the sequenced clones have been siblings?  Have you repeated the experiment and also recovered the same 2kb part of chromosome 13q?

I had ever successfully sequenced 2 carriers before using the same method, in that case, I got over 15 clones with the correct insert. I didnt care about the rest clones at that time. after this problem happened, I checked the sequences of rest failed clones, and I found the same sequence as what I got lately. So this might not be an accidental thing.

Here's the point -- if your 2kb part of chromosome 13q inserts were amplicons, then they would have the primer sequences in them.  Since they do not, and it looks like the banding pattern you're seeing is a result of amplification based on your latest photo, we have to assume that the band your seeing is a mixture of amplicons (in vast excess) and co-migrating genomic fragments of the same size.  During this experimet, just due to chance, you caught a genomic fragment when cloning, and what you sequenced were all siblings.

Strange things happen when cloning -- I once recovered a segment 100% identical to a tomato gene while cloning a band from E. coli -- no idea how it happened, but I repeated the experiment, and got the segment I wanted.

There are two major issues here -- your primers are not specific enough for what you are trying to do, as they produce way too many bands, and you're relying on blunt-end cloning.  Primer specificity could be improved by re-design or changing the PCR conditions and/or cycling parameters.  Blunt-end cloning could be avoided by attaching restriction enzyme sites to the ends of your primers, or, in this case, perhaps a better solution would be to try TA cloning.  Use a Hi-Fi Taq-based enzyme, and clone into a TA TOPO vector.  Either of these methods will insure that you can only successfully clone PCR products, and take the genomic fragments out of the equation.

I redesigned a primer set which is 25bp in length and no improvement at all. this primer set works well in the cell line. It also seems not the problem of primers. I think at least we can draw a conclusion here that copy number of proviral genome in carriers is extremely low. because there shouldnot be a big difference between genomic DNA of the cell line (maybe T cells) and PBMC.
Anyway, I should consider the next step, maybe TA cloning. I used Herculase II before, which is 6 times higher than taq in fidelity. I searched Hi-Fi taq-based enzyme and found easy-A from stratagene, which is also 6 times more faithful than taq. that is a choice. Do you have any recommendations?

Thank you very much, Homebrew. I really appreciate your help and I have learned a lot from you. wish i could solve this problem.

[HomeBrew]

gyma, on Apr 3 2010, 06:25 AM, said:

Do you have any recommendations?

We use the Platinum PCR SuperMix High Fidelity from Invitrogen.  It too has 6X fidelity versus Taq alone.

There is another experiment I'd try.  Prepare four PCR reactions -- one with both primers, one with just the forward primer, one with just the reverse primer, and one with neither primer added -- and cycle them all.  See what banding pattern is produced in each case.

What are your PCR cycling conditions?  Can you post your primer sequences?

[gyma]

HomeBrew, on Apr 3 2010, 10:10 PM, said:

gyma, on Apr 3 2010, 06:25 AM, said:

Do you have any recommendations?

We use the Platinum PCR SuperMix High Fidelity from Invitrogen.  It too has 6X fidelity versus Taq alone.

There is another experiment I'd try.  Prepare four PCR reactions -- one with both primers, one with just the forward primer, one with just the reverse primer, and one with neither primer added -- and cycle them all.  See what banding pattern is produced in each case.
thats a good idea. I will try that later.Thanks.

What are your PCR cycling conditions?  Can you post your primer sequences?

PCR conditions:
95c, 5min;
30cycles: 95c,30s;62c,30s;72c,1min (30s/kb for the enzyme I used);
72c, 10 min.
old primer set: F:CAGCGGTTACAAAACCGACA; R: AGCAGTTCAGGAGGTGCCGA.
new primer set:F: GCTCTACTCCTCCTCGTCATATTGTT; R: AGCTCGACCTGAGAGGAGACTTACC.

[gyma]

I discussed this with other people. some said there might be contamination in my PCR system. but if so, why water never showed a band?
I did what homebrew had recommended in his last reply and I found only REVERSE primer could get similar band patterns, which means this primer served as both Forward and Reverse ones. later I found the primer sequence in the genomic sequence that I cloned and it indeed was both REVERSE primer sequence at each end. The reason why I didnt find that before is that this sequence is not 100% same, and before I just copied the primer sequence and then searched for homology. because of the several-base-difference, I got nothing at all. Now I learned a lesson here, even if the last base of a primer doesnt mach, it is still possible to amplify a product. only the yield wouldnot be so good.
Anyway, it turns out the primers were not good, especially the reverse one, which caused all the nonspecific bands. However, what should I do next? I redesigned a new primer set and that didnt work neither. I suspect that the proviral load might be at an undetectable level of PCR, or at least it is equally hard to amplify proviral sequence or nonspecific human genomic products. Am I right?
Thank you guys who ever helped me in this post, especially homebrew. I really appreciate your help. of course I will appreciate more if you keep helping me

[HomeBrew]

gyma, on Apr 5 2010, 11:54 AM, said:

I did what homebrew had recommended in his last reply and I found only REVERSE primer could get similar band patterns, which means this primer served as both Forward and Reverse ones.

I suspected one of the primers alone was going to produce your bands -- nothing else made sense.

gyma, on Apr 5 2010, 11:54 AM, said:

Now I learned a lesson here, even if the last base of a primer doesnt mach, it is still possible to amplify a product. only the yield wouldnot be so good.

The primer itself is incorporated into the PCR product, so it must match the primer -- it IS the primer.  If there are base differences between your insert sequence and your expected primer sequence, there are two possibilities -- the primers were synthesized sloppily, or the base calling in the primer region of your insert sequence is incorrect.

gyma, on Apr 5 2010, 11:54 AM, said:

However, what should I do next?

Must you amplify the whole 2kb region in proviral genome?  Is your goal to detect the presence of these proviral genomes or to clone them in their entirety?  If detection is sufficient, move your reverse primer such that it pairs with your forward primer to amplify a 1.5 kb, 1.0 kb, or 500 bp piece -- I'm sure that at some point, the primer pair will be specific to your proviral genome and not amplify any spurious bands...

[gyma]

Thanks a lot, homebrew.
my goal is to sequence some genes in the proviral genome, so I have to amplify the genes first.
Now I am going to optimize the PCR conditions and try to get the specific product. because the conditions optimized from the cell line didnt work in carriers, so I have to use the precious sample now. seems no other choice.