PCR principles and practices


Important parameters in the PCR:


  I. Template DNA quantity (complexity determines ng) and quality (average length)

      While people typically measure DNA quantity in ng, the relevant unit is actually moles, i.e.,

      how many copies of the sequence that will anneal with your primers are present.

      Thus, the amount of DNA in ng that you need to add is a function of its complexity.  

      In theory, a single molecule of DNA can be used in PCR but normally people use between

      1000 and 100,000 molecules for eukaryotic nuclear DNA.


       Example for sorghum (genome size = 760 Mb):

            760 Mb = 7.6 x  108  bp x  649 daltons/bp = ~5 x 1011 grams/mole

            20 ng = 2 x 10-8 grams/  5 x 1011 grams/mole = 1 x 10-18 mole

            1 x 10-18 mole x 6 x  1023 molecules/mole = 6 x 104 molecules


       Example for a 5 kb plasmid clone:

            5 kb = 5 x  103  bp x  649 daltons/bp = 3.2 x 106 grams/mole

            1 pg = 1 x 10-12 grams/ 3.2 x 106 grams/mole = 3.2 x 10-18 mole

            3.2 x 10-18 mole x 6 x  1023 molecules/mole = 2 x 106 molecules


      Template quality and PCR product size both affect the amount of DNA you need to add. If

      your DNA is very high MW, and/or your product length is short (e.g., an SSR), you can use

      less DNA because a higher fraction of the molecules will contain the annealing sites for both

      the forward and reverse primer. If your DNA is degraded and you want to amplify a large

      product, it may not work, but the same DNA may be fine for amplifying SSRs.


  II. Tm of primers

       The melting temperature of your primers should be similar and should be as high as possible,

       within reason, in order to increase specificity. “Within reason” means that you don’t want the

      Tm of the primer to be higher than the reaction temperature of Taq polymerase (72 C). In

      practice, I usually aim to have my primer Tms around 66 C, and this is usually possible with a

      primer length of 22 or 23 bases. These primers can frequently be used in 2-step PCR (see



To calculate Tm for duplex DNA <50 bp:


               Add 2deg. C for each A or T

               Add 4deg. C for each G or C


      The annealing temperature you use in your PCR should be a function of the Tm of your

      primers, and it should not be much lower unless you have designed the primer from

      heterologous sequence, in which case you may want to do a gradient PCR (see below).




III. Mg concentration

       Standard Mg++ concentration is 2 mM, but sometimes you may find that the concentration

      needs to be raised (rarely lowered) to get a PCR to work. Raising Mg lowers specificity, and is

      roughly comparable to lowering the annealing temperature. It may cause multiple bands to

      appear (or, occasionally, disappear). In my experience, most troubleshooting is more easily

      accomplished by playing with other variables (temperature, DNA quality or amount).


IV. Length of expected product

      The length of the extension step (72 C) should be a function of the length of the product you

      are  trying to amplify. A general guideline is 1 minute/kb of product length, but in fact this is

      more than is needed, particularly if you are doing 3-step (i.e., conventional) PCR, as extension

      will take place during the annealing step and during the ramp time. Taq polymerase is a very

      fast and very processive enzyme.


Optimizing your reaction, saving time


The original PCR programs typically used 1 minute denaturation and  annealing  steps. This is way more time than is needed. Also, if your product will be much less than 1 kb, a 1 minute extension is also too long.


When possible, use 2-step PCR:

            1. Initial denaturation, 2-3 min at 94 C

            2. Denaturation, 15 sec at 94 C

            3. Simultaneous annealing and extension, x minutes at 68 C

                x depends on product length, 1 min/kb

            4. Return to step 2 for 29-39 additional cycles.


When the Tm of your primers is below 65 C, or when 2-step PCR does not yield a product, use

3-step PCR:

            1. Initial denaturation, 2-3 min at 94 C

            2. Denaturation, 15 sec at 94 C

            3. Annealing, 15 sec at x C

                x depends on Tm, should be as high as possible, often ≥ Tm

            4. Extension, x sec

                x depends on product length, ≤ 1 min/kb

            5. Return to step 2 for 29-39 additional cycles.


Most PCR programs set up in this way should be complete in well under 2 hours. This saves you time and frees up the machine for other users. Shorter times also prolong the activity of your Taq in the later cycles.

I usually let the thermocycler block heat up to at least 80 C before I put my tubes in. This is not a true hot-start but I think it may improve specificity.


When your program is finished, remove tubes as promptly as possible and CANCEL YOUR PROGRAM. Do not leave an open block set on 4 C!!

Gradient PCR

When you are trying out heterologous primers, you can run a set of PCRs with different annealing temperatures all in the same block. This saves time and does not tie up multiple blocks.


Long PCR

Long PCR requires special enzymes and concentrations of reaction components. Often it is advisable to, after ten cycles, add 10 seconds to the extension time of every consecutive cycle, as the Taq polymerase loses activity and takes more time to make the full length product.


Nested PCR

This is a very reliable way to get a better PCR product when results are marginal for a number of reasons. Design a new F or R primer (F’ or R’) internal to the original, compatible with the partner, and use a tiny amount of the primary PCR product as template in the PCR with F + R’ (or F’ + R, as the case may be).

Frequently this will

-eliminate extra bands if the first PCR was messy

-produce a robust band where the first PCR was weak or even invisible

This method also saves on genomic DNA. For example, you could do a small-scale primary PCR of a 2 kb product using a modest amount of DNA. It might not work very well, but from that primary PCR you can amplify two overlapping 1.2 kb products for sequencing (using primers F + R’ and F’ + R). 

Remember that you need only a very little of the primary product in the nested PCR because it is very low complexity (see above). I often dilute the primary PCR 1:100 and add 1 uL to the nested PCR.  Sometimes this is too much. Also, you often can cut your cycles back to 25-30 for the nested PCR. You have to determine this empirically.


Finally, a few words about enzymes

Enzymes are stored at -20oC in a non-frost free freezer, typically in 50% glycerol. The tubes should never be allowed to reach room temperature and gloves should be worn when handling or you must handle the tube very carefully and keep your fingers away from the opening. Always use a new, clean pipet tip every time you use a stock tube of enzyme.


Before you open a new tube of enzyme, spin it briefly first as there is often enzyme in the cap. This is particularly true for temperature-sensitive enzymes (i.e., all except Taq), that may be put in ice: enzyme in the cap does not stay cold. Spin tubes as necessary to keep enzyme at bottom.


Avoid trying to measure out minute quantities of enzyme, as the 50% glycerol storage buffer makes this impossible. When pipetting enzyme from a stock tube, place the end of the tip just far enough into the enzyme to get what you need, do not plunge the tip way down into the solution, as the outside of the tip will become covered with enzyme and your measurement will be off.

Whenever possible, make a “cocktail” of enzyme, buffer, water, etc., and aliquot this as appropriate. (DO NOT add enzyme to unbuffered water, you can denature it. Mix water and buffer first, place on ice, then add enzyme.) The volume of the enzyme should be less than 1/10 of the final volume of the reaction mixture, as too much glycerol can interfere with enzyme activity.


Enzymes are expensive and perishable. Use them carefully!!