Tips for increasing your PCR specificity (decrease nonspecific product formation)

When it comes to PCR, the thing I typically care most about is specificity. I want my sequence of interest to be copied (amplified) and NOT anything else. So when I go to the imager and see multiple bands in my gel lane, indicative of the presence of other, “non-specific” products I know there’s optimization ahead! Here are some things I turn to…
blog: bit.ly/pcrspecificity
There are some things you can test out that include changes to the primer, but first I want to talk about what I usually try first - just changing the PCR reaction conditions without having to design and buy new primers…
First and foremost, try raising the anneal temperature. This is the temperature during the step when the primers bind to the template after each cycle (after the melt step that unzips the double strands of the previous cycle’s products (amplicons)). The higher the temperature, the more energy molecules have to squirm. And the more they squirm, the harder it is for them to stay stuck to anything. So it’s harder for the primers to stick to the template at higher temperatures. The more attracted a molecule is to something, the more energy (heat) is required to unstick. So, by using an anneal temperature that is too high to stay stick to those nonspecific sites, but low enough that the molecules will stay stuck to the target site, you can increase your specificity.
*key point: the higher the anneal temperature the greater the specificity (assuming your target sequence is “the best site” on the template which we will come back to when discussing primer design).
*increase the anneal temperature to decrease production of specific bands
How do you know how high to set the anneal temperature? You might need to test out a range (temperature gradient settings on the PCR machine are great for this) but a good place to start is the Tm. This is the “melting temperature”, the temperature at which, on average, half of the primer copies are predicted to be bound. Typically you set the anneal temp a couple degrees below this as a starting point (but try using out Tm prediction tools that take into account the polymerase type and reaction conditions to get a more realistic Tm).
*The higher the Tm of the primers, the higher you can take your anneal temp.
And note that I said primerS. You have two of them. And you will be limited by the lower one. So it’s best to design your primers to be close to one another in Tm (within a couple degrees) so you can optimize them together and not risk going too high for one and too low for another.
If your Tms are really high, you can skip the anneal step altogether and go straight from melt to extend (this is referred to as 2-step PCR).
What makes a Tm high? Things that make it stickier and/or increase the number of binding interactions between the primer and the template. So,
*the longer the primer (assuming it binds the whole length) - if you need to raise the Tm of one primer to match the other, simply lengthening it a few bases typically works
*the higher the Tm & the higher the GC content, the higher the Tm
So why not just chock your primer full of Gs & Cs? This runs the risk of causing weird secondary structures (folded shapes), primers binding to themselves or the other primer (primer diners), etc. And those things will take primer out of the search for template (reduce your effective concentration) and can even cause short, primer-amplified-primer products to form.
I recommend using a primer checking program (I use AmplifiX) to check your primers for predicted problems like that when you design them. If you can’t avoid them by simply shortening or lengthening your primer, try moving your primer binding site if possible.
While we’re on the topic of primer design… be sure to check that your primer isn’t predicted to have strong partial-complementarity (part of it matches) to other potential sites. The larger the template is, the more other sites there are which you need to worry about. So start with as pure a sample as possible (eg make sure your plasmid isn’t contaminated with E. coli genomic DNA from the cells you grew it in).
If you’re doing something where you want to amplify from that big thing, the chances of those off-target “targets” occurring by chance are higher. But thankfully you can make lower the chance by increasing the primer length (similarly to why you increase your password length to make it harder to guess).
Now let’s get back to the anneal temperature…
finished in comments

[@thebumblingbiochemist]

We said that higher is better in terms of specificity, which is true. But too high is worse in terms of efficiency and yield. Remember that above the Tm, most copies of the primer won’t be bound. So there’s a sort of best-of-both-worlds strategy you can try called Touchdown PCR. It starts with a cycle or two at a high anneal temp, so you give the most specific products an (exponential) head start and then you decrease the anneal temp in subsequent cycles so you can increase yield - with copies coming from those original good copies that will swamp out any subsequent bad ones.
Another temperature-based strategy to try is using HotStart Polymerases. These have something that makes them inactive (such as an antibody bound to them) until you raise the temperature in that initial melt step and that blocking thing falls off and/or gets inactivated, etc. more on this here:   bit.ly/hotstartPCR & czcams.com/video/mshTkT8YKTM/video.html
But the purpose of it is that it prevents the polymerase from doing any copying while you are setting up your reaction and you have lower temperatures, potentially higher concentrations of primers, etc.
Which brings me to another thing you can optimize to improve specificity…
*decrease template and/or primer concentration to reduce risk of nonspecific product formation
The more wrong sites there are for the primer to find, the more likely you’ll get bad products. And the more primer copies there are, the more chances there are for mispriming to occur. High primer concentrations also raise the risk of primer dimer-related problems.
Finally (though probably not last in terms of order of testing since it’s simple) you can try adding an additive like DMSO which can decrease secondary structure in your template, help the template strands melt, and thus make it easier for the primer copies to find the target you want. This is especially important if the template region containing that target is GC-rich. There are also specific buffers you can use with GC-rich sequences.
I recommend optimizing for specificity and then, if you need a higher yield, perform multiple PCR reactions and pool them together during purification - you can prepare a scaled-up reaction in one tube but then split it down to the volume you optimized at for the actual running (i.e. don't try to run it on the large volume). More on this here: blog: bit.ly/pcryield ; CZcams: czcams.com/video/B3AhFvhYNk0/video.html  
Hope that helps!
Promega has a nice guide on optimizing your PCR: www.promega.com/resources/guides/nucleic-acid-analysis/pcr-amplification/#general-considerations-for-pcr-optimization-6f575242-fe99-4fcb-88c5-b205ad7becc7
And here are some links to other posts of mine
PCR playlist: czcams.com/play/PLUWsCDtjESrHbBVTt6wqWBGGbd0Q9pbRA.html
more on PCR: bit.ly/pcrtrain & czcams.com/video/GZSLfECgW3Q/video.html  
more on nucleic acid spin columns: bit.ly/spincolumns & czcams.com/video/fz2OpjxQKKM/video.html  
 
more on agarose gel electrophoresis: bit.ly/agarosegelcompare & czcams.com/video/vbuxf3rcMxg/video.html  
 
more practical lab tips: bit.ly/lab_tricks_page & czcams.com/play/PLUWsCDtjESrFEAWZCRKJL7sMc6a_KgfLU.html
  
more about all sorts of things: #365DaysOfScience All (with topics listed) 👉 bit.ly/2OllAB0 or search blog: thebumblingbiochemist.com

[@haribhaskaran8607]

Great video for troubleshooting your PCR! Thank you

[@thebumblingbiochemist]

Thanks! Happy to help

[@guacamole8554]

Thank you 🙏

[@cezreycor]

I recommend a smaller glove size. other than that, great vid! you're amazing! :) thanks!

[@thebumblingbiochemist]

Thanks so much - so glad you found it helpful!