PCR : Learning the Cheats - Less is More

OK, I had a meeting with my adviser earlier today that prompted me to make this post. Unlike some postdocs, I actually have a pretty decent relationship with my PI : he treats me fairly and has good insight. Today was one of those days "good insight" days where I left his office thinking, "Hmmm, that's true. I bet lots of people are confused about this." What the heck am I talking about? Well, something as cool (read:mundane) and sexy (read:boring) as PCR template concentration and optimization.

How many times have you unwrapped the casing on a PCR kit, unfolded the protocol (!) and it says something like: Template DNA, use between 10 pg and 1 ug. OK, sure... that's a range of what?

A million fold.

Why not just say "Put some DNA into dah tube". Sheesh. This can be confusing for students who are uninitiated to PCR. I bet it is even confusing to some postdocs. I think this is why so many people have to be trained on _how_ to do PCR. Of course, because there are so many variations of how to use PCR at the bench, you often end having to optimize each PCR reaction to suit each specific template.

A million fold. It just make the little scientist in me laugh. This huge range is a major over-simplification of the real issue : stoichiometry. There's a Golden Rule (at least in my book) about PCR that everyone should remember: Less is More. The less template you add, the more specific your amplification will be. The less chance there is of mispriming. Of course, you are limited by the sensitivity of your system (Phusion ftw!), but in general you want to find the sweet spot and add as little of your DNA as possible.

For example, the mass to unit ratio for a sample of human genomic DNA is, for example, extremely high. For every template unit (i.e. individual amplifiable region, lets say ~100 nt) there is a tremendous amount of mass associated with it (each DNA strand may be millions of base pairs in length). For sample such as these, Less is More. Start with about 200ng, and do 2-fold serial dilutions down to about the 1 pg range. Don't forget your no-template control (NTC) - always a must do.

For artificial DNA constructs, plasmids, and even DNA oligos, the mass to unit ratio is much lower. That is, the ratio of the total mass of DNA compared to the mass of the region you want to amplify is much lower. Is this situation, the same rule applies: Less is More. Start with even less mass (~ 10 or even 1 ng) than the above example, and then optimize using a 2-fold serial dilution down to the femptamolar range.

The extreme example of mass to unit ratio would be in a SELEX experiment we are seeking to amplify a very small amplicon (85 nt) from a very small template (85 nt). The mass to unit ratio is 1. For SELEX, the first round of the experiment is important as you want to make sure you have as much coverage of your aptamer library as possible. For a 20-mer random aptamer, the math of this works out that we need about 1.8 pmols of template for complete coverage of our SELEX library (about a trillion individual molecules to start). This turns out to be only about 4.9 ng of ssDNA. Furthermore, for subsequent rounds of SELEX - you need even less because presumably your library of aptamers is being reduced (which is the goal of SELEX). In fact, after the initial round of SELEX, you only need a few picograms, or even femtograms, of DNA for your PCR amplification - and even this represents PLENTY of template.( On a side note: For anyone who is interested, I started a running series on SELEX a few weeks ago.)

I'm not sure what I was aiming for here; but maybe some of you will benefit at the bench from this little PCR cheat : Less is More.

Oh, and why the Nintendo Controller? Well... you know why.

EDIT AFTER POST: There a follow up to this post that can be found here.