Hi Ikram, I guess your main problem is amplification of genomic DNA, which often is. But I don't see why you should complicate your life with such long primers, normally between 30-35 nt are sufficient, considering 15-20 nt of homology and 12-15 nt on the 3' side as proper primer. The design of this 3' side is very important, this is the one annealing to your genomic DNA, therefore it should contain some 60% GC and should have a strong 3' ending, which means 3 GC pairing within the last 5 nt, possibly the last two at the 3' end being CC or GG (this stabilizes the pairing where the Taq will start polymerizing, but will also allow you to work at higher annealing temperatures and reduce unspecific binding of the primers that reduces efficiency of your PCR reaction).

Secondly, often amplification of genomic DNA requires higher primer concentration. Standard PCR conditions use 0.25 uM primer concentration, but you can actually go up to 1 uM when amplifying genomic DNA... this is essentially due to the fact that there is so much sequence in the genome that your primers end up stochastically binding everywhere, although unspecifically, but are titrated away from the specific binding equilibrium in the specific site.

Thirdly, the longer the fragment you are trying to amplify, the more difficult, although there are specific Taq polymerases to try to deal with this (Expand from Roche, Q5 from NEB, Phusion from NEB or Thermo Scientific, and may others). In general try not to go for very long pieces, let's say not longer than 2 kb. It turns out to be quicker and easier to produce a long fragment by putting together several shorter pieces, either by fusion PCR or the Gibson Assembly itself...

Fourth, GC content of the region you need to amplify will also affect reaction efficiency. Above 40% GC content it might become very difficult in standard conditions. You can think of using the special mixes for high GC content provided by the companies or try yourself adding to the basic PCR buffer 1-5% DMSO or up to 1M betaine or combinations of the two.

Another point to take into account is the number of cycles in your PCR. For proofreading reasons suppliers normally suggest to keep the cycle number low, around 25 cycles. Anyway if you have no bands at all you can try increasing the cycle number up to 35. An additional trick is to re-amplify a first reaction, better if with nested primers. The idea is to design a first set of primers external to the primers you will need to produce your fragment for the Gibson Assembly. Run a first PCR with the first set, which might not give any visible bands. Then take between 0.1 and 1 ul of this reaction and use it as template for a new PCR with your Gibson Assembly primers. You clearly increase the chance of introducing mutations... but you might end up with a product. You'll have to sequence anyway your clones at the end of the cloning.

Finally there are lots of possibilities with step-up, step-down, temperature gradient, and time gradient PCR protocols... You might want to specifically search for them!

Good luck

Hi Ikram, I guess your main problem is amplification of genomic DNA, which often is. But I don't see why you should complicate your life with such long primers, normally between 30-35 nt are sufficient, considering 15-20 nt of homology and 12-15 nt on the 3' side as proper primer. The design of this 3' side is very important, this is the one annealing to your genomic DNA, therefore it should contain some 60% GC and should have a strong 3' ending, which means 3 GC pairing within the last 5 nt, possibly the last two at the 3' end being CC or GG (this stabilizes the pairing where the Taq will start polymerizing, but will also allow you to work at higher annealing temperatures and reduce unspecific binding of the primers that reduces efficiency of your PCR reaction).

Secondly, often amplification of genomic DNA requires higher primer concentration. Standard PCR conditions use 0.25 uM primer concentration, but you can actually go up to 1 uM when amplifying genomic DNA... this is essentially due to the fact that there is so much sequence in the genome that your primers end up stochastically binding everywhere, although unspecifically, but are titrated away from the specific binding equilibrium in the specific site.

Thirdly, the longer the fragment you are trying to amplify, the more difficult, although there are specific Taq polymerases to try to deal with this (Expand from Roche, Q5 from NEB, Phusion from NEB or Thermo Scientific, and may others). In general try not to go for very long pieces, let's say not longer than 2 kb. It turns out to be quicker and easier to produce a long fragment by putting together several shorter pieces, either by fusion PCR or the Gibson Assembly itself...

Fourth, GC content of the region you need to amplify will also affect reaction efficiency. Above 40% GC content it might become very difficult in standard conditions. You can think of using the special mixes for high GC content provided by the companies or try yourself adding to the basic PCR buffer 1-5% DMSO or up to 1M betaine or combinations of the two.

Another point to take into account is the number of cycles in your PCR. For proofreading reasons suppliers normally suggest to keep the cycle number low, around 25 cycles. Anyway if you have no bands at all you can try increasing the cycle number up to 35. An additional trick is to re-amplify a first reaction, better if with nested primers. The idea is to design a first set of primers external to the primers you will need to produce your fragment for the Gibson Assembly. Run a first PCR with the first set, which might not give any visible bands. Then take between 0.1 and 1 ul of this reaction and use it as template for a new PCR with your Gibson Assembly primers. You clearly increase the chance of introducing mutations... but you might end up with a product. You'll have to sequence anyway your clones at the end of the cloning.

Finally there are lots of possibilities with step-up, step-down, temperature gradient, and time gradient PCR protocols... You might want to specifically search for them!

Good luck

What are the expected sizes for your PCR products? Which polymerase are you using? What are your cycling conditions? Do you have a positive control for your PCR?

I would agree with many of the things Pietro said, but also add:

PCR is sensitive to the amount of template you have in the reaction, and with genomic DNA you have so much of it that is nothing to do with what you are doing. For (high quality) genomic DNA I get good results with around 60ng per reaction (I keep a diluted stock at 60ng/ul to make setting up the PCR easier)

A ~45-mer primer should be long enough, 20-25nt homology for the fragments and 15-20nt for template annealing. That's not to say that your current primers can't work!

With LARGE primers I have had luck in the past using Phusion with annealing temperatures between 65C and 72C, +/- DMSO depending on the template. I tend to do a gradient from 58C to 72C.

As Pietro also says, there are many many parameters you can change in PCR, but in my experience you can usually get something to amplify with a temperature gradient +/-DMSO in the unmodified buffer (I mean no Mg) supplied with the enzyme (provided you follow the basic primer design rules). After that you can optimise further if you need to.

Dear Dr Cohen

The expected size of my products is 3.3 Kb. I am using Taq Polymerase. The annealing temperatures for the primers are high in the range of of 69-77. Rest I follow the standard PCR protocol.

Dear Dr Boyl

Thank you very much for your suggestions. If I design the primers as you have suggested i.e. 15-25 nts of homology and 12-15 nts of proper primer with over 55% of GC then should I consider the annealing temperature of only this  part or the whole primer (complete 35-40 nts). Also, the final concentration of my primers in the reaction is 0.4uM.

The larger the PCR product, the more difficult the amplification from a genomic DNA template. While feasible, amplifying a 3.3kb fragment is not an easy task. Also using Taq polymerase, which has relatively low fidelity is not recommended. A polymerase with higher proofreading activity such as Pfu or Phusion would be better. Third, your annealing temperature should be below the extension temperature, no matter what the predicted Tm is. I would recommend you try the annealing at 55C or 58C instead as an initial starting point. If you have a thermocycler that will let you try a gradient of annealing temperatures, then make use of this feature to find an optimal Tm.

Lastly, genomic DNA preps can have lots of RNA and protein contamination that can inhibit PCR and also mislead your estimates of nucleic acid concentration. Unless you have a positive control showing that your genomic DNA functions as a good PCR template, you might benefit from cleaning up your gDNA prep with a Proteinase K treatment, followed by phenol chloroform extraction, and RNase treatment.

Hi Ikram, correct, however you decide to design your primers (and as you see different strategies are possible, so as Michael says, yours might be fine and before you design new ones you should make sure they really don't work) you should calculate or, better, estimate the annealing temperature based on the number of bases that actually anneal to your genomic DNA target. In general this is sufficient for the whole PCR reaction, although in some cases, if you end up with too much unspecific, you can think of raising the temperature after 5-10 cycles to account for the fact that the template after these cycles will include also the 5' homology region in the majority of the molecules.

Secondly, I think that whatever are the primers you are presently using, the annealing temperature range that you mention in your reply to Daniel Cohen is too high. Michael is correct in all he suggests, you should start with annealing temperature around 58°C and then raise it if you obtain unspecific (or do a gradient PCR if your machine allows it, with a 3°C step, to spare time).

I also strongly support Michael's suggestion to keep genomic DNA amount low. You can try a titration also with that (so, different amounts), but be sure you don't use more than 200 ng, otherwise you strongly inhibit the Taq polymerase... Quantifying genomic DNA is not always simple, depending on the extraction method: if you don't remove the RNA efficiently and you end up degrading it substantially you will not be able to quantify it properly.

I think you should also verify better which Taq polymerase you are using among the many and many variants (engineered or from different archea species) proposed by companies. There are differences, indeed, in their processivity and efficiency...

Dear Dr Cohen and Dr Boyl

Thank you very much for your suggestions. Just before checking your latest comments I set up another PCR, this time using a High Fidelity Taq polymerase (Expand HiFi enzyme-Plus) from Roche. I choose 3 different annealing temperatures - 55, 57 and 59. Also, the concentration of my genomic DNA is 45.6 ng/ul and I am using 1ul of this sample. Lets see what happens.

Thank you

Good luck, Ikram. If this latest round does not succeed, I would recommend trying touchdown PCR cycling parameters. Another challenge that you may need to consider, depending on the source of your genomic DNA, is the potential influence of SNPs on annealing of your oligos. Mismatches between your oligo and template may ensue if SNPs cause sequence differences between the actual sequence present in your gDNA and the predicted sequence present in a reference sequence used to design your oligo.

Dear Dr Cohan and Dr Boyl

The reactions have started working now. I think the main trouble was with annealing temperatures as I was not considering only the actual annealing sequences. Also, I increased the primer concentration. I really appreciate your helpful suggestions and now I will set up the assembly reactions.

Thank you very much.

Ikram