Inbiochemistry, apolymerase is anenzyme (EC 2.7.7.6/7/19/48/49) that synthesizes long chains ofpolymers ornucleic acids.DNA polymerase andRNA polymerase are used to assembleDNA andRNA molecules, respectively, by copying a DNA template strand usingbase-pairing interactions or half ladder replication.
A DNA polymerase from thethermophilic bacterium,Thermus aquaticus (Taq) (PDB1BGX, EC 2.7.7.7), is used in thepolymerase chain reaction, an important technique ofmolecular biology.
A polymerase may be template-dependent or template-independent.Poly-A-polymerase is an example of template independent polymerase.Terminal deoxynucleotidyl transferase is also known to have template independent and template dependent activities.
| Produces DNA | Produces RNA | |
|---|---|---|
| Template is DNA | DNA-dependent DNA polymerase (DdDp) or commonDNA polymerases | DNA-dependent RNA polymerase (DdRp) or commonRNA polymerases |
| Template is RNA | RNA-dependent DNA polymerase (RdDp) orReverse transcriptase (RT) | RNA-dependent RNA polymerase (RdRp) orRNA-replicase |
Polymerases are generally split into two superfamilies, the "right hand" fold (InterPro: IPR043502) and the "double psibeta barrel" (often simply "double-barrel") fold. The former is seen in almost all DNA polymerases and almost all viral single-subunit polymerases; they are marked by a conserved "palm" domain.[2] The latter is seen in all multi-subunit RNA polymerases, in cRdRP, and in "family D" DNA polymerases found in archaea.[3][4] The "X" family represented byDNA polymerase beta has only a vague "palm" shape, and is sometimes considered a different superfamily (InterPro: IPR043519).[5]
Primases generally don't fall into either category. Bacterial primases usually have the Toprim domain, and are related totopoisomerases and mitochondrial helicasetwinkle.[6] Archae and eukaryotic primases form an unrelated AEP family, possibly related to the polymerase palm. Both families nevertheless associate to the same set of helicases.[7]
Scientists have modified the activity of nucleic acid polymerases in many ways, from rational design to directed evolution, to achieve changes from incremental tweaks like higher speed/accuracy/thermostability or major shifts such as conversion of template and product types.
All known natural reverse transcriptases evovled from an ancestor that has no proofreading ability, causing a low fidelity. In 2016, scientists successfully useddirected evolution to modify the proofreadingThermococcus kodakarensis DNA-directed DNA polymerase into what they call areverse transcribing xenotranscriptase (RTX). This new enzyme is able to copy from and proofread with RNAand DNA templates. It is expected to improve the accuracy in RNA sequencing and other forms of RT-PCR.[8] It was commercialized some time before April 2018.[9]
In 2022, selective mutagenesis converted aKod DNA polymerase into one that produces α-l-threofuranosyl nucleic acid orthreose nucleic acid (TNA).[10] This result has been improved in 2024 and 2025 using HR-accelerated directed evolution, yielding several enzymes with near-natural speed and fidelity.[11]
In 2025, scientists used directed evolution, accelerated by homologous recombination (HR), to change a DNA polymerase into an RNA polymerase. It is able to perform transcription quickly (3 nt/s) and accurately (>99%). It is also a somewhat "universal" polymerase, being also capable of RNA-directed DNA production (reverse transcription) and chimeric DNA–RNA amplification.[12]
