Articles

PCR Fundamentals for Cloning

Polymerase fidelity, cloning primer tail design, two-stage annealing temperatures, and cycle limits

Molecular Cloning

PCR Fundamentals for Cloning

Years ago, if you wanted an insert, you had to cut it out of a plasmid using whichever restriction sites happened to be in the right places. PCR removed this constraint. Gibson Assembly, Golden Gate, In-Fusion and most modern methods now assume you will make your insert by amplification.

This changes what a primer is for. With a custom primer, you amplify your target and, in the same reaction, add an overlap, an overhang, a restriction site or a promoter sequence to the end of the product.

It also changes what counts as success. A diagnostic PCR answers one question: is the sequence there or not. A cloning PCR has to produce a building block with exactly the right ends and no mutations, because whatever comes out of the tube is what ends up in your construct. This guide covers polymerase choice, primer design, and the cycling conditions that decide whether your product is usable.

At a glance

  • Polymerase: Taq for screening only. A proofreading enzyme for anything you will express.
  • Primers: The 3' end anneals to the template. The 5' tail adds what your cloning method needs, and does not anneal at first.
  • Watch for: Calculating the melting temperature across the whole primer instead of the template-binding part.

Choosing a polymerase

Your polymerase decides two things at once: how many mistakes end up in your insert, and what the ends of the product look like.

Taq, for screening

Taq was the original thermostable PCR enzyme. It is fast, cheap and robust, and it has no 3'-to-5' exonuclease domain, which means no proofreading. Its error rate is often quoted as roughly one mutation per few thousand bases of product.

It also tends to add a single adenine to the 3' end of each new strand, leaving a one-base A overhang. This is a nuisance for most methods and the basis of TA cloning.

Do not use Taq to amplify a gene for an expression construct. The arithmetic is against you: a 2 kb insert amplified with Taq will often carry a mutation, and you will find it at the sequencing stage after you have already screened colonies. Keep Taq for diagnostic colony PCRs. You are only asking whether a band is the right size, so an occasional wrong base changes nothing, and a cheap master mix saves the high-fidelity enzyme for the reaction where fidelity decides the result.

High-fidelity polymerases, for cloning

Q5, Phusion, PrimeSTAR and similar enzymes have a 3'-to-5' exonuclease domain. When the polymerase adds a mismatched base, the distortion it creates moves the DNA into the exonuclease site, the wrong base is removed, and synthesis continues. Error rates are reported up to a hundred times lower than Taq's.

The proofreading activity also trims back overhanging bases, so these enzymes leave blunt ends. This is what blunt-end cloning and some multi-part assemblies need.

They cost more per reaction, and for a cloning insert this is not a place to save money. Sequence the product anyway: proofreading lowers the error rate; it does not remove it.

Primer design for cloning

A diagnostic primer is around twenty bases that match your target. A cloning primer has two sections doing two different jobs.

The 3' section anneals to your template and does the priming. The 5' tail does not bind the template at all, at least not at the start. It carries whatever your method needs:

  • Restriction cloning: a few spacer bases, then the restriction site. The spacers are there because an enzyme cuts poorly when its site sits at the very end of a fragment.
  • Gibson or HiFi: 20 to 40 bases matching the adjacent fragment or the vector.
  • In-Fusion: 15 bases of homology, at the absolute end of the fragment.
  • Golden Gate: the inward-facing Type IIS recognition site and your chosen four-base overhang.
  • Directional TOPO: a CACC sequence at the 5' end of the forward primer (see TOPO Cloning).

These tails make the primer long, and long primers behave differently. They can fold into hairpins, or pair with each other instead of the template.

Pairing at the 3' ends is the form that costs you the reaction. A polymerase needs an annealed, free 3'-OH group to start, so two primers joined at their 3' ends make a usable priming site. The enzyme extends them into short, stable, double-stranded products that amplify efficiently, consume your reagents and outcompete the product you wanted. Interactions involving the 5' ends do much less damage, because there is no anchored 3' end to extend, and they melt apart during denaturation.

Plan it first

Complex cloning primers are hard to get right by eye. One miscalculated overlap, or a missing spacer in front of a restriction site, shows up two steps later as an assembly that does not work.

Before you order oligonucleotides, build the construct on screen. In Photo51, you can lay out your template, map the primer tails onto the fragment ends and simulate the PCR, so you see the product with its added ends and can check that it will fit the vector the way you intend. The aim is to confirm that the fragment you are about to order will assemble into the construct you have in mind (see our deep dive on PCR Primer Design).

Troubleshooting

A long primer gives no product at all

Symptom: You designed a 50-base primer for a Gibson Assembly, calculated the melting temperature of the whole primer, set the annealing temperature to match, and the gel is empty.

Cause: The melting temperature was calculated across a sequence that is not annealing. In the first few cycles, only the 3' section of the primer binds the template. The 5' tail hangs free. A Tm that includes the tail is far higher than the temperature at which the primer can actually bind, so your annealing step is too hot for anything to happen.

Fix: Calculate the Tm from the template-binding sequence only, and set the first three to five cycles to match it. Once the full primer has been copied into the product, you can raise the annealing temperature for the remaining cycles to suit the full-length primer. This two-stage program is commonly used for long cloning primers.

A smear instead of a band

Symptom: Your product runs as a smear across a range of sizes rather than one crisp band.

Cause: The annealing temperature is too low, so primers bind despite several mismatches, in places you did not intend.

Fix: Raise the annealing temperature to make the binding more selective. A gradient PCR across 55 °C to 65 °C will find the point where you get a single specific product, and it costs you one run rather than three guesses.

Nothing amplifies from a GC-rich template

Symptom: A target above about 65% GC gives no product, or the polymerase appears to stall part way through.

Cause: GC-rich single-stranded DNA folds into stable structures, including hairpins and G-quadruplexes, that the polymerase cannot read through.

Fix: Add DMSO or betaine rather than changing the enzyme. Both are standard additives that destabilize these structures, DMSO by lowering the dielectric constant of the solution and betaine by changing how the template is hydrated. Which one works better is usually a matter of trying them.

Point mutations in an otherwise correct clone

Symptom: Your PCR worked, the assembly worked, the colony grew, and sequencing shows scattered point mutations in the insert.

Cause: Too many cycles. Even a proofreading polymerase makes occasional mistakes, and an error made in an early cycle is copied into every molecule descended from it, so the mutated fraction of the product rises with each extra cycle.

Fix: Keep cloning PCRs between 25 and 30 cycles. This gives you plenty of DNA for an assembly and keeps the mutation load low. If 30 cycles does not give you enough product, the problem is the reaction conditions, and running 40 cycles will not fix it.