PCR Primer Design and Reaction Optimisation

Standard PCR Protocol

Ed Rybicki

Copyright January 1994, February 2001




Recommended Reagent Concentrations:

  • Primers: 0.2 - 1.0 uM
  • Nucleotides: 50 - 200 uM EACH dNTP
  • Dimethyl sulphoxide (DMSO): 0 - 10% (v/v)
  • Taq polymerase: 0.5 - 1.0 Units/50ul rxn

Target DNA: 1 ng - 1 ug (NB: higher concn for total genomic DNA; lower for plasmid / purified DNA / virus DNA target)

Buffer: use proprietary or home-made 10x rxn mix; eg: Cetus, Promega. This should contain: minimum of 1.5mM Mg2+, usually some detergent, perhaps some gelatin or BSA. Promega now supply 25mM MgCl2, to allow user-specified [Mg2+] for reaction optimisation with different combinations of primers and targets.

MAKE POOLED MASTER MIX OF REAGENTS IN ABSENCE OF DNA using DNA-free pipette, then dispense to individual tubes (using DNA-free pipette), and add DNA to individual reactions USING PLUGGED TIPS.

OVERLAY REACTIONS WITH 50UL OF HIGH-QUALITY LIQUID PARAFFIN OR MINERAL OIL to ensure no evaporation occurs: this changes reactant concentrations. NOTE: latest wisdom has it one can use VASELINE - this also allows "HOT START" PCR.



USE PLUGGED PIPETTE TIPS: prevents aerosol contamination of pipettes.

Use of detergents is recommended only for Taq from Promega (up to 0.1% v/v, Triton X-100 or Tween-20). DMSO apparently allows better denaturation of longer target sequences (>1kb) and more product.


Remember sample volume should not exceed 1/10th reaction volume, and sample DNA/NTP/primer concentrations should not be too high as otherwise all available Mg2+ is chelated out of solution and enzyme reactivity is adversely affected. Any increase in dNTPs over 200uM means [Mg2+] should be re-optimised.


Low primer, target, Taq, and nucleotide concentrations are to be favoured as these generally ensure cleaner product and lower background, perhaps at the cost of detection sensitivity.


Recommended Reaction Conditions:

Initial Conditions:

Initial denaturation at start: 92 - 97oC for 3 - 5 min. If you denature at 97oC, denature sample only; add rest of mix after reaction colls to annealing temperature (prevents premature denaturation of enzyme).

Initial annealing temperature: as high as feasible for 3 min (eg: 50 - 75oC). Stringent initial conditions mean less non-specific product, especially when amplifying from eukaryotic genomic DNA.

Initial elongation temperature: 72oC for 3 - 5 min. This allows complete elongation of product on rare templates.

(see also here)


Temperature Cycling:

  • 92 - 94oC for 30 - 60 sec (denature)
  • 37 - 72oC for 30 - 60 sec (anneal)
  • 72oC for 30 - 60 sec (elongate) (60 sec per kb target sequence length)
  • 25 - 35 cycles only (otherwise enzyme decay causes artifacts)
  • 72oC for 5 min at end to allow complete elongation of all product DNA


"Quickie" PCR is quite feasible: eg, [94oC 30 sec / 45oC 30 sec / 72oC 30 sec] x 30, for short products (200 - 500 bp).


DON'T RUN TOO MANY CYCLES: if you don't see a band with 30 cycles you probably won't after 40; rather take an aliquot from the reaction mix and re-PCR with fresh reagents.  See here.


"Hot Start" PCR:

In certain circumstances one wishes to avoid mixing primers and target DNA at low temperatures in the presence of Taq polymerase: Taq pol is almost as efficient as Klenow pol at 37 oC; consequently, if primers mis-anneal at low temperature prior to initial template denaturation, "non-specific" amplification may occur. This may be avoided by only adding enzyme after the initial denaturation, before the reaction cools to the chosen annealing temperature. This is most conveniently done by putting wax "gems"TM into the reaction tube after addition of everything except enzyme, then putting enzyme on top of the gem: the wax melts when the temperature reaches +/-80oC, and the enzyme mixes with the rest of the reaction mix while the molten wax floats on top and seals the mix, taking the place of mineral oil. Information is that "gems" may be substituted by VaselineTM.


Asymmetric PCR for ssDNA Production:

Simply use a 100:1 molar ratio of the two primers (eg: primer 1 at 0.5uM, primer 2 at 0.005uM). This allows production of mainly ssDNA of the sense of the more abundant primer, which is useful for sequencing purposes or making ssDNA probes.


Detecting Products:

Take 1/10th - 1/3rd of the reaction mix CAREFULLY from under the oil or from under the Vaseline or solidified wax, using a micropipette with plugged tip, IN AN AREA AWAY FROM YOUR PCR PREPARATION AREA! 

Mix this with some gel loading buffer(1:1 - 1:5 mix:loading buffer): this is TBE containing 10 - 20% glycerol or sucrose and a dash of bromophenol blue (BPB) tracking dye. 

Load 5 - 30ul of sample into wells of 0.8 - 3.0% submarine agarose gel made up in TBE, preferably containing 50ng/ml ethidium bromide

Run at 80 -120 volts (not too slow or small products diffuse; not too fast or bands smear) until BPB reaches end of gel (large products) or 2/3 down gel (small products). Use DNA markers going from 2kb down to 100 bp or less (recommend BM PCR markers). 

View on UV light box at 254 - 300 nm, photo 1 - 5 sec.


Small products are best seen on 3% agarose gels that have been run fast (eg: 100 volts), with BPB run to ½ - 2/3 down the gel. It is best to include EthBr in the gel AND in the gel buffer , as post-electrophoresis staining can result in band smearing due to diffusion, and if there is no EthBr in the buffer the dye runs backwards out of the gel, and smaller bands are stripped of dye and are not visible.

NUSIEVE TM gel (FMC Corp) can also be used for small products - better resolution than agarose.

Polyacrylamide gels can be silver stained by simple protocols for extreme sensitivity of detection.

Gels can be blotted directly after soaking in 0.5M NaOH / 1.5M NaCl for 10-20 min: "dry blotting" works well (eg: gel is over- and under-layered with paper towel stacks and pressed; bands transfer up and down), as does classic "Southern" blotting. Bands blotted in this way are already covalently fixed onto nylon membranes, and simply need a rinse in 5xSSPE before prehybridisation.

The example shown is of detection of Human papillomavirus type 16 (HPV-16) DNA amplified from cervical biopsy samples ( Williamson A-L, Rybicki EP (1991) Detection of genital human papillomaviruses by polymerase chain reaction amplification with degenerate nested primers. J Med Virol 33: 165-171).  The left panel is a photo of an EthBR-stained 2% agarose gel; the right is an autoradiograph of a Southern blot probed with 32P-labelled HPV-16 DNA.  Note how much more sensitive blotting is, and how much more specific the detection is.


Labelling PCR Products with Digoxigenin

PCR products may be very conveniently labelled with digoxigenin-11-dUTP (Boehringer-Mannheim) by incorporating the reagent to 10-35% final effective dTTP concentration in a nucleotide mix of final concentration 50-100uM dNTPs (Emanual, 1991; Nucleic Acids Res 19: 2790). This allows substitution to a known extent of probes of exactly defined length, which in turn allows exactly defined bybridisation conditions. It is also the most effective means of labelling PCR products, as it is potentially unsafe and VERY expensive to attempt to do similarly with 32P-dNTPs, and nick-translation or random primed label incorporation are unsuitable because the templates are often too small for efficient labelling.

Make a DIG-dNTP mix for PCR as follows:


  • Dig-11-dUTP 350 uM
  • dTTP 650 uM
  • dATP 1 mM
  • dCTP 1 mM
  • dGTP 1 mM

For each 50 ul of probe synthesized, a 1/10 dilution is made of the DIG-nucleotide mix when added to the other reagents as described above. The products may be analyzed by agarose gel electrophoresis - NOTE: PRODUCTS ARE LARGER THAN NON-SUBSTITUTED PRODUCT - and detected directly on blots immunologically. Probes can be used as 5-10 ul aliquots directly from PCR product mixes, mixed with hybridisation mix and denatured. Probes can be re-used up to 10 times, stored frozen in between experiments and boiled to denature.


Cleaning PCR Products

  • Getting rid of mineral oil: simply add 50ul of chloroform to the reaction vial, vortex and centrifuge briefly, and remove the "hanging droplet" of AQUEOUS solution with a micopipette.
  • Getting rid of wax or Vaseline: simply "spear" wax gem and remove; do as for oil or bottom-puncture tube and blow out aqueous drop for Vaseline.
  • Cleaning-up DNA: 3 options
    • a protocol which gives DNA that is clean enough for sequencing is the following: increase volume to 100ul with water, add 10M ammonium acetate soln. to 0.2M final concentration (1/5th volume), add equal volume of isopropanol (propan-2-ol), leave on bench 5 min, centrifuge 20 min at 15 000 rpm, remove liquid using pipette, resuspend in 100ul water or TE, repeat precipitation.
    • Simply do a phenol-CHCl3 extraction (add 20ul phenol to aqueous phase, vortex, add 50ul CHCl3, vortex, centrifuge, remove UPPER aqueous phase, repeat CHCl3 extraction).
    • Make aqueous phase up to 400ul, and spin through Millipore Ultrafree-MC NMWL 30 000 cartridges (at 6000 rpm in microcentrifuge), wash through with 2x400ul water, collect +/-20ul sample: this is pure enough for sequencing.


Product is clean enough for restriction digest immediately after reaction; however, phenol-chloroform clean-up is recommended as a minimum.


Sequencing PCR Products:

This is best done using ssDNA generated by asymmetric PCR, and the "limiting" primer for sequencing. However, efficient sequencing of dsDNA generated by normal PCR is possible using the modification to the SequenaseTM protocol published by Bachmann et al. (1990) (NAR 18: 1309). CLEAN DNA is resuspended in sequencing buffer containing 0.5% Nonidet P-40 and 0.5% Tween-20 and sequencing primer, denatured by heating to 95oC for 5 min, snap-cooled on wet ice, and sequenced by the "close-to-primer" protocol (eg: dilute extension mixes).


Cloning PCR Products

T-A Cloning Strategy: Taq and other polymerases seem to have a terminal transferase activity which results in the non-templated addition of a single nucleotide to the 3'-ends of PCR products. In the presence of all 4 dNTPs, dA is preferentially added; however, use of a single dNTP in a reaction mix results in (relatively inefficient) addition of that nucleotide. This complicates cloning, as the supposedly blunt-ended PCR product often is not, and blunt-ended cloning protocols often do not work or are very inefficient. This can be remedied by incubation of PCR products with T4 DNA pol or Klenow pol, which "polishes" the ends due to a 3'->5' exonuclease activity (Lui and Schwartz, 1992; BioTechniques, 20: 28-30). However, this terminal transferase activity is also the basis of a clever cloning strategy: this uses Taq pol to add a single dT to the 3'-ends of a blunt-cut cloning vector such as pUC or pBluescriptTM, and simple ligation of the PCR product into the now "sticky-ended" plasmid (Marchuk et al., 1990; NAR 19: 1156).

Incorporation of Restriction Sites in Primers: Although this may be rendered simple by incorporating the same or different restriction sites at the 5'-ends of PCR primers - which allows generation of sticky ends and straightforward cloning into appropriate vectors - these should have AT LEAST two additional bases 5' to the recognition sequence to ensure that the enzymes will in fact recognise the sequence - and it is often found that even when this is done, the efficiency of cutting of fresh product is next to zero. This can sometimes be remedied by incubating fresh product with Proteinase K (to digest off tightly-attached Taq pol), but often is not. A solution to the problem is to use the "Klenow-Kinase-Ligase" (KKL) method: this involves "polishing" products with Klenow, kinasing them to get 5'-phosphorylation (NB: OLIGONUCLEOTIDE PRIMERS NORMALLY HAVE NO 5'-PHOSPHATES!!!), ligating the fragments together to get concatemers, then restricting these with the appropriate restriction enzymes to generate the sticky-ended fragments suitable for cloning (Lorens, 1991; PCR Methods and Applications, 1: 140-141).