Protocol Online logo
Top : Forum Archives: : Molecular Biology

Trouble PCRing a long (7.5 kb) fragment - (Mar/02/2006 )


I can't seem to be able to PCR a 7.5kb fragment from genomic fly DNA. I'm using Invitrogen's Platinum Pfx polymerase since it apparently can amplify up to 12 kb and is hi-fi. The primers I designed initially have a Tm of ~66 C, but since I also had to add restriction sites onto the 5` end, this raised the Tm to ~76 C. Here are the thermocycler settings I used:

initial denature 2 min 94 C
denature 15s @ 94 C
anneal 30 s @ 60 C
extend 8 min @ 68 C
-->>30 rounds
Pause @ 4 C

This protocol was recommended by the Pfx manual, but it didn't work. Instead I got a tiny amount of a bunch of random fragments between 500 bp - 2500 bp. Increasing the denature and annealing times to 1 minute didn't help. I figured the annealing temperature was too low, so I was getting non-specific bindings. So, I tried this instead:

initial denature 5 min @ 94 C
denature 30s @94 C
anneal 1 min @ 60 C
extend 8 min @ 68 C
-->>4 rounds
denature 30s @ 94 C
extend 8 min @ 68 C
-->>27 rounds
Pause @ 4C

The rationale behind this was that the first few rounds would allow for the primers to bind to the genomic template (which obviously don't have the 5' restriction sites I added), but once I started getting some product, I increased the annealing temp to match the Tm of the primers with the extra restriction sites added. The actual annealing step was eliminated since the annealing temperature is roughly the extension temperature (recommended by the Pfx manual) (this paragraph is confusing, I'm sure... but I hope I'm making sense!)

After doing this, I still got the same result... a bunch of random junk no bigger than ~2.5 kb... certainly nothing close to 7.5 kb. Using a different, cleaner template also did not help. I also checked the concentration of the template and it was no more than 25 ng/ul, so I wasn't saturating the reaction. The primers I designed had no significant dimerization or repeats or anything like that. Moreover, I used primers for a 1.2 kb fragment as a control and got positive results from that every time.

So in short... can anyone suggest anything? I'm really stuck!!


in my opinion, expand high fidelity PCR system ( from roche) worked much better than pfx when i amplified 2.5kb from human genomic DNA. the latter didn't work at all. but Roche one did amplified.
you could try it. good luck.


the initial denaturing temperature ..I would reduce it to 1 min and try
also I have obtained better results amplifying 6 kB fragments form genomic DNA using Roche as well Elongase from Invitrogen

genomic DNA PCR can be tricky and trouble some..dont lose heart...try some more conditions ..maybe some other primers ..increasing Mg conecntration ...also u can order a library instead of using genomic DNA


one thing that helps when sequencing a large template (maybe it will work here) is to preheat treat the template. take the template in water, heat treat at 96C for 1-3 min, return to room temperature then add the rest of the reaction components and cycle.


you can do a PCR in pre heated block, and start initial denaturation for 1'.
And you may add DMSO to 2%max for reducing non specific bands.


I had a similar problem, trying to amplify a 2.1 kb template in a plasmid. But finally it worked. I used Pwo-polymerase and increased the extension time to 4 min. (referring to the manual: 2 min. per 1kb), further I added extra MgSO4 and I am running temperature gradient PCRs all the time, in order to find the optimal annealing temperature for the primers (for example: I experience shorter products too, but only at lower annealing temperatures).

I also observed that a "hot start" PCR works better when using proofreading enzymes:
I preheat the thermocycler, after adding polymerase, I put the reaction mixture on ice and from the ice directly to the preheated cycler.

In my opinion, you should increase the extension time up to 15 min. try also adding extra Mg and try a temperature gradient, maybe you should increase the denaturing step to 1 min.

here is my protocol which I use succesful for amplifying my 2.1 kb template:

preheating the cycler to 95°
initial denaturing: 5 min @ 95°
denaturation 1 min @ 95°
annealing 30 s @ 55° ±5° (gradient)
extension 4 min @ 70°
--> 30 cycles
final extension 10 min @ 70°
cooling @ 4°

I don't know if it works with genomic DNA , but try as many conditions as possible!

good luck!


I'd try some annealing temperatures lower than the ones you have tried. 55C and 58, e.g. I'd also try a longer denaturing at 94, both initially (say make it 5 minutes) and in each cycle (say 30 - 60 sec). The extension temperature might also be too short. If none of these work, I would suggest redesign of the primers, which is the usual problem. Since primers are so cheap these days it does not pay to optimize conditions for poor primers. I'd suggest redesign using Primer3. If there is a very low GC region in your expected product, you may need to lower the extension temperature.