Primer design help - secondary structures - Please help me troubleshoot? (May/13/2008 )
Hello there! Anyone care to help a noob out? 
Basically there’s an element within of a protein that we are interested in and in order to study that, I’ve been asked to design primers (include restriction sites) to generate a variety of DNA fragments to use as inserts into vectors. Think of it as somewhat of an elimination process where certain regions of the cDNA are excluded and the consequences observed when that particular fragment is digested and ligated into a vector.
My understanding of the primer designing process with regards to this is that it is a fairly ‘strict’ process. For example if I am interested in creating a fragment that will eventually be translated into residues 20-300 of my protein, I will need a forward primer that looks like this:
5’ - NNNN – Restriction site – region of overlap of approx 15 bp starting with the first nucleotide of residue #20 - 3’
So I came up with the sequences to my primers, and I put them through the Net Primer to check for any secondary structures etc… To my horror they all came up with multiple putative hairpin structures, self dimerization, cross dimerization and at that point my hair was starting to turn grey.
Now it’s not like I input my DNA sequence into a primer designing program and let them spit out the best possible primers matches, since I need them to generate specific regions of the protein I don’t have the flexibility to choose another part of the DNA that will give me primers with less secondary structures/lower Tm.
Are there any solutions to this? Or should I just go ahead and deal with it? OR is my understanding of this primer designing business is completely WHACK? (i.e. if I’m interested in aa residues 20-300, I don’t necessarily need to have my 15bp overlap starting exactly on the first nucleotide of residue #20)
And if you have read this far, thanks! Heh.
Good morning, 
It's true that just about any DNA sequences will, by chance, have some amount of self-complementarity or ability to dimerize. This is particularly true if you are trying to add restriction sites to a PCR fragment: by their nature, RE sites have the ability to dimerize with themselves.
Happily, this is rarely a problem in PCR because you can adjust the stringency of the annealing conditions. I'd bet that pretty much all of the secondary structures or dimers that Net Primer gave you involve a relatively few number of bases binding to each other. These structures and dimers won't be able to form at higher temperatures because they will all have a Tm that is much lower than that of your PCR priming site. This is why it's important to carefully choose your PCR annealing conditions: too high a Tm and your primer won't bind, too low and you can have non-specific primer binding interactions.
When setting up a PCR such as you are, I generally run the first few cycles using a Tm equal to the Tm of the 15 bp complementary region (although I normally use 20 bp to increase specificity) and then run the remaining cycles at a Tm of the full-length primer (including NNNN-restriction site) to make sure the PCR is as specific as possible.
Hopefully this helps somewhat. Also, as a hint, if are planning to do express this PCR fragment you might want to think about adding a Kozak sequence between the restriction site and the start codoon.
Have a great day!
Ginger
15 bp of complementarity is not really adequate. Go to 18-24, depending on the GC content of the region. Try to make the 3' end GC rich, but without a long string of GC. The problem sequences have the 3' end of a primer matching well to a region of a primer (itself or the other) such that it can extend. Many of the patterns shown by the IDT tool, e.g., cannot be extended, and are not important. Many others have weak association and can be ignored. You have a problem when there is an excellent 5+ bp match at the 3' end of a primer to another region of a primer (itself or another) which can be extended. This is worse if other parts of the primer also match well.
Thanks for your input!
I have done plenty of PCRs before, and the common annealing temperature is often between 55 to 60 degrees Celcius... And I understand that the annealing temperature should be a couple degrees off the Tm of the primers. However, when I analysed my primer sequences using Net Primer, some of the Tms were in the 70-ish degrees region! Is that still okay because it just seems ridiculously high to me :S
I spoke to my supervisor about including the Kozak sequence, however, he said that it is already within the vector that we are using (apparently it was cloned into the vector together with the FLAG tag). I was just wondering, did I manage to get the primer designing concept right at all???
Thanks
Does this mean that I don't have to worry about most of the secondary structures unless there is more than 5bp of complementary nucleotides?
I get nervous when I see 3 bp matches at the very 3' end along with say half a dozen additional matches further 5', where the 3' end can be extended.
I aim for an annealing temp 5-10 degrees below the theoretical Tm. Also, I don't bother using the algorithms from primer design apps, I just go with the old "4 degrees for G or C, 2 degrees for A or T" rule of thumb. And as others have said, I would suggest you try 20-24 nt for the complementary sequence.
No problems whatsoever, i always design my primers with Tm greater than 70C. Why not? I never understood why people don't do this, it can only increase specificity by being able to use a higher annealing temperature.
Well check to make sure it is there, pretty important! Does it need to be next to the start codon though? That's where we usually put it.
I usually design ~22-25nt primers, with sequences a little more GC rich on the 5' end and AT rich on the 3' end. I use software (Primer Select or VectorNTI) to check for hairpins and dimers. I try to avoid any hairpin with a negative delta G, and dimers with anything below ~ -3 delta G. (Redesign if needed!) My TM's can range anywhere from ~60°C - ~75° or 80°, depending what I'm working on. But I try to make sure the primers are within 5° of each other, and usually set my annealing temp 5-10° below my lowest TM.
Well check to make sure it is there, pretty important! Does it need to be next to the start codon though? That's where we usually put it.
That’s what I thought too… My forward primers don’t even include the start codon. He said I could put it in but it doesn’t make much of a difference if I don’t since “it’s already in the vector”.
I have included stop codon for my reverse primers though.
I’m actually more concerned whether if I have gotten the whole primer designing concept right (e.g. what are the necessary elements to include) and being able to tell the different types of primers apart (for PCR, cloning, sequencing etc).