Avoiding mutations during PCR cloning

PCR cloning copies an insert with a polymerase and then traps one of those copies inside a plasmid. The copy can be the right length and the wrong sequence. Avoiding mutations means choosing a proofreading class when the insert will be trusted, stopping the cycles once you have enough product, cloning more than one colony, and sequencing the whole insert. A single pretty band is not that sequence. How the cycle itself works is in how polymerase chain reaction works. How a checked insert becomes an archived colony is in plasmid cloning from insert to colony.

The photograph shows a gloved hand closing the lid of a benchtop thermal cycler on a strip of tubes. The programme behind that lid, especially the cycle count, is part of the mutation problem. Enzyme classes are in the molecular biology catalogue. The Thermo Fisher PCR overview is one manufacturer's map of polymerase classes. It is not a protocol to copy onto a different enzyme.

Where the errors come from

A Taq-like polymerase synthesises DNA and does not carry a 3-prime to 5-prime exonuclease that would remove a wrong base. When it inserts a mismatch, later cycles copy the mistake as if it were truth. Early errors become a large share of the molecules. Late errors stay rare. That is why cycle number matters as well as enzyme class. Proofreading polymerases do have that exonuclease. They leave fewer mistakes. They do not leave none, and a long insert still deserves a sequence even when the tube says high fidelity.

Damage and oddities sit beside misincorporation. A depurinated template, too much enzyme, or a tired nucleotide mix can raise the mess. So can a primer that already contains a wrong base: every molecule inherits it, and every colony will match each other and miss the design. That pattern is the clue. Shared mutations point at primer or template. Scattered mutations point at the polymerase during the run.

The insert you clone is one molecule. The band on the gel is a population. Colony screening by size draws from the population and still cannot see a substitution. Insert orientation can be correct on a mutant. Both screens stay necessary. Neither replaces the read.

A path from a bad protein back to the tube

Start with the phenotype people actually notice. The protein is the wrong size, inactive, or different from the translation you wrote down. Pull the plasmid you expressed and compare its full insert sequence to the design, not to your memory of the colony PCR. If you only sequenced the junctions, sequence the middle before you change media. A mutation in the centre of a two-kilobase gene is invisible from the ends of a short read.

Then ask whether other colonies from the same ligation share the change. Open two more. If all three carry the same substitution, the primer or the template already had it. Order a corrected primer, or pick a different template stock, and amplify again. If the three colonies disagree, you cloned individual polymerase errors. Pick further colonies from the same plate only if the plate still exists and the errors are sparse. Otherwise repeat the PCR with a proofreading enzyme and fewer cycles, and sequence several new colonies. Do not site-direct a fix onto a clone you have not fully read. A second mutation may be sitting downstream.

If the sequence matches the design and the protein still does not, the mutation hypothesis is retired. Frame, promoter, host and induction are the next questions. This page stops at the sequence. It does not pretend every failed induction is a Taq error.

Addgene's PCR notes remind you to keep a no-template control so a contaminating amplicon is not the "mutation." A contaminant is the wrong sequence too. NCBI GenBank is a place the reference coding sequence may be, so you know what "the design" was. Addgene's protocol collection is a wider set of cloning notes. Cycle numbers still belong to the enzyme card and to how much template you really added.

What to change in the amplification

Use a proofreading polymerase when the product will be cloned and expressed or used as a reference. Keep Taq-like enzymes for screens where a band is the answer and the DNA will not become a plasmid you trust. If you must clone a Taq product, sequence more colonies, not fewer, and expect an added base at the ends if you are using an A-overhang method. That extra base is a cloning feature. An extra base in the middle is a mutation. Do not confuse them.

Limit cycles. When the template is a pure plasmid or a generous cDNA, a shorter programme often yields enough insert. Match the extension time to the enzyme class. A classical Taq-like planning figure is on the order of one minute per kilobase near 72 Celsius. Proofreading enzymes publish other rates. Follow the enzyme. Do not "make sure" by doubling both the cycles and the extension until the band is a smear. A smear cloned as one band is a mixture.

Clone from a reaction that had a no-template control and a single band you can explain. Gel-purify if primer-dimer or a second species is present. Ligate, transform, and pick several colonies. Sequence the whole insert of at least two that passed a size screen. Agreeing sequences that match the design are the ones you may keep. A pretty chromatogram of the wrong colony is still the wrong colony.

Do not trust the band, and do not trust one colony

The troubleshooting trap is aesthetic. The gel photograph looks clean, the colony PCR looks clean, and the first sequencing file is only the junctions. The protein then refuses to match. The missing evidence was the middle of the insert in a second colony. Build that evidence before the induction, not after a week of buffer changes. If the insert is long, several primers must cover it. A mutation between two reads is unread, not absent.

If every attempt with PCR yields a different error and the gene is repetitive or long, stop amplifying it. A synthetic insert designed as the sequence you want removes the polymerase lottery. That is a sourcing decision, not a confession that cloning failed morally.

Safety

A mutated insert can be more hazardous than the one you designed, or simply nonsense. Your institution's biosafety rules apply to what you actually cloned, once you know it, and to the template you amplified. This note is research practice. It does not validate a clinical sequence and it does not approve an edited gene for release. Hot lids and ultraviolet gel boxes are the everyday physical hazards. Close the lid, as the photograph shows a gloved hand doing, and do not stare into an unshielded UV source.

A cycler that rebooted mid-programme

After a power cut a thermal cycler may resume, restart, or sit at one temperature while the display looks finished. The cycle count you intended is then a guess, and a guess is a bad input to a fidelity plan. Discard that tube for cloning. Run the amplification again when the block is stable, and record the cycle number you set. Do not clone a pretty band from a run you cannot describe. Two cyclers in the same warm room can also deliver different effective times if their blocks differ. The enzyme card's temperature is the target. The instrument log is the evidence.

What to say in an enquiry

Name the polymerase class you need, proofreading or not, the insert length, the template, and the fact that the product will be cloned and fully sequenced. Ask for the enzyme's own extension guidance rather than assuming a generic minute per kilobase. If you would rather start from a designed sequence than from a PCR population, use the custom gene synthesis enquiry reference as an enquiry. Reads that cover the whole insert can be framed with the Sanger sequencing enquiry reference. Neither page means EVRINTH runs synthesis or sequencing. Send the enzyme and primer requirement through the quote request and ask whether a quotation is possible.