Science Wiki

Primers are short single-stranded nucleic acid polymers that provide the mandatory starting point for DNA synthesis. Because replicative DNA polymerases can add nucleotides only to a pre-existing 3′-hydroxyl (3′-OH) group, they cannot initiate a new polynucleotide chain de novo. A primer is therefore annealed to the single-stranded template to supply this reactive 3′-OH nucleophile. Primers are composed of RNA in cellular chromosomal replication, of DNA or a protein-linked hydroxyl residue in certain viruses and plasmids, and of chemically synthesized DNA oligonucleotides in in vitro techniques such as the polymerase chain reaction (PCR) and Sanger sequencing.

The requirement for a primer is shared by replicative DNA polymerases and by many reverse transcriptases. It is not shared by RNA polymerases, which can start synthesis without a primer, and a few specialized enzymes can also start DNA synthesis de novo. The composition and the source of the primer differ widely among biological systems.

Two-Metal-Ion Phosphoryl Transfer

A primer provides a 3′-hydroxyl group that is positioned in the active site of a polymerase, paired to the template. In the two-metal-ion mechanism that is common to polymerases, divalent cations (in cells, Mg²⁺) bind the carboxylate residues of the active site. One metal ion lowers the p_K__a of the 3′-hydroxyl group, so that a 3′-oxygen nucleophile forms. It attacks the innermost (α) phosphate of the incoming deoxynucleoside triphosphate (dNTP), forms a phosphodiester bond, and releases pyrophosphate. The primer is thus extended in the 5′ to 3′ direction:

3'-Terminal Base-Pairing Specificity

The polymerase recognizes the primer-template junction by features of the duplex, including the minor groove, and not by sequence. For this reason, a primer must be paired to the template at its 3′ end. A mismatch at the 3′ terminus lowers the rate of extension by orders of magnitude, which is the basis of allele-specific PCR and of the mismatch discrimination used in genotyping. It also means that a primer that is not annealed correctly is extended poorly.

RNA primers made by primase

In the chromosomal replication of cells, primers are RNA. A primase binds to the single-stranded DNA that is exposed by the helicase, and it synthesizes a short RNA primer. Primases are not evolutionarily related in all organisms. The bacterial primase DnaG belongs to the TOPRIM fold, whereas the primases of archaea and eukaryotes belong to the archaeo-eukaryotic primase (AEP) superfamily. In eukaryotes, the primase is a heterodimer of PRIM1 (catalytic) and PRIM2 that forms a complex with DNA polymerase α (Pol _α_-primase).

The primers are short. Bacterial primers are usually about 10 to 12 nucleotides long, and eukaryotic primase makes RNA of about 7 to 10 nucleotides. In eukaryotes, the Pol _α_-primase complex then moves the primer to the polymerase subunit, which extends it with about 20 deoxynucleotides, so that the final RNA-DNA primer is about 30 nucleotides long. Pol α lacks proofreading activity. The DNA part of the primer has a relatively high error rate, and it is later removed in part by the strand displacement and flap processing of Okazaki fragment maturation.

The two strands are made in different ways. On the leading strand, a primer is needed only once at the origin, and later synthesis is continuous. On the lagging strand, a new primer is needed for each Okazaki fragment. In E. coli, Okazaki fragments are about 1,000 to 2,000 nucleotides long, so a genome of 4.6 megabases needs thousands of primers. In eukaryotes, the fragments are about 100 to 200 nucleotides, so the human genome needs millions of primers.

In bacteria, DnaG interacts with the DnaB helicase and has weak affinity for the template, which limits primer synthesis to when it is needed. Primase prefers certain trinucleotide sequences (5′-CTG-3′ in E. coli, where the primer begins with the complementary pppAG). In eukaryotes, Pol _α_-primase is attached to the replisome through its interaction with the CMG helicase and with the factor And-1 (Ctf4), and the Ctf4 trimer links Pol α to the helicase.

Why RNA?

The reason for the use of RNA as a primer is not settled. A common proposal is that primers are a hazard because a polymerase that starts a chain without a primer has no proofreading, and that the initiation is inaccurate. An RNA primer can be recognized as different from DNA by the repair and processing machinery, which then marks it for removal and replacement by DNA that is of higher quality. The RNA-DNA hybrid also differs in structure: it has an A-form helix, and RNase H and flap endonucleases can recognize it. Another argument is that RNA polymerases, unlike DNA polymerases, can start chains without a primer, so primase can use this ancestral ability. These explanations are not mutually exclusive, and they are partly hypotheses about the evolutionary origin.

Primer removal

Before the lagging strand is complete, the RNA primers must be removed and the gaps filled. The pathway differs among organisms:

• In bacteria, DNA polymerase I removes the RNA by its 5′ to 3′ exonuclease activity while it synthesizes DNA behind it, which is called nick translation. DNA ligase then seals the nick. RNase H can also remove part of the RNA.
• In eukaryotes, Pol δ extends an upstream fragment and displaces the downstream primer into a flap. The flap is cut by flap endonuclease 1 (FEN1), or by Dna2 if the flap is long. RNase H2 can remove the RNA, with FEN1 removing the last ribonucleotide. DNA ligase I seals the nick.

At the ends of linear chromosomes, the removal of the terminal RNA primer on the lagging strand leaves a gap that cannot be filled, because no upstream 3′-OH is available. This is the "end replication problem", which leads to the shortening of telomeres with each division unless it is compensated by telomerase. Telomerase is itself a reverse transcriptase that uses its own RNA template and the 3′ end of the chromosome as a primer.

Priming from a Nick or a 3′ End

Some replication systems use the 3′-OH of the DNA itself. In rolling circle replication, an initiator endonuclease (Rep protein, relaxase, or a related enzyme) cuts one strand of a double-stranded circular DNA at a specific site. The cut strand has a free 3′-OH, and a DNA polymerase uses it to start synthesis, which displaces the 5′ end. The initiator remains covalently attached to the 5′ phosphate through a tyrosine. This mechanism is used by many plasmids and by single-stranded DNA phages such as ΦX174, and the initiator enzyme there is the viral A protein, not a "relaxase" in the narrow sense that is used for conjugative plasmids (where the relaxase initiates transfer, and a rolling circle mechanism then replicates the transferred strand).

Other examples of DNA priming include:

• The hairpin ends of the parvovirus genome, which are folded into a duplex that provides a 3′-OH, so that the genome primes its own replication.
• The 3′-OH of a broken DNA end during break-induced replication and recombination-dependent replication, in which a strand that has invaded a homologous template is a primer.
• The 3′ overhang of a telomere in the case of telomerase, as above.
• The annealed 3′ end of a repeated sequence in some mechanisms of template switching.
• The 3′-OH at a nick or gap in DNA repair, for example in base excision repair, where a polymerase fills a gap that is bounded by an upstream 3′-OH.

Protein Priming

In protein-primed replication, a terminal protein supplies the hydroxyl group. This is used by the linear genomes of the bacteriophages _Φ_29 and PRD1, of adenoviruses, and of some linear plasmids and mitochondrial genomes. A serine, threonine, or tyrosine residue of the terminal protein (a serine in adenovirus, and a serine in _Φ_29 terminal protein) attacks the _α_-phosphate of the first dNTP and forms a covalent protein-nucleotide bond. The polymerase then continues synthesis. In _Φ_29 and adenoviruses, the terminal protein forms a complex with the DNA polymerase, and the product remains covalently attached to the 5′ end of the new strand. This provides a solution to the end replication problem of linear genomes without RNA primers.

Priming by tRNA and Other RNAs in Reverse Transcription

Retroviruses use the 3′ end of a host transfer RNA as the primer for reverse transcription. HIV-1 uses tRNA^Lys,3, which anneals to the primer-binding site next to the 5′ end of the genome, and murine leukemia virus uses tRNA^Pro. During reverse transcription, short purine-rich RNA fragments of the genome that remain after RNase H digestion (the polypurine tracts) serve as primers for the second strand. Non-LTR retrotransposons, such as LINE-1, use target-primed reverse transcription, in which the element's endonuclease nicks genomic DNA and the exposed 3′-OH primes cDNA synthesis directly on the RNA template.

De Novo Initiation by DNA Polymerases and Primase-Polymerases

Some enzymes start DNA synthesis without a primer. The human protein PrimPol, a member of the AEP superfamily, was identified in 2013 and has both primase and polymerase activity. It can make short DNA primers de novo and re-prime DNA synthesis downstream of lesions that stall the fork, in the nucleus and in mitochondria. Primases in the AEP superfamily from bacteria and archaea can also make DNA or RNA. Other proteins, such as some DNA polymerases ε, δ, and θ under some conditions, have been reported to initiate or extend with unusual substrates, but these are exceptions that are not part of the normal pathway.

Comparative Overview of Biological and Synthetic Priming Modalities

Synthesis of oligonucleotide primers

Synthetic primers are made by oligonucleotide synthesis with the phosphoramidite method, in which nucleotides are added one at a time to a growing chain on a solid support in the 3′ to 5′ direction (the reverse of the direction of enzymatic synthesis). Each cycle has four steps: removal of the 5′ dimethoxytrityl protecting group (detritylation), coupling of the next phosphoramidite that is activated by a tetrazole derivative, capping of unreacted chains, and oxidation of the phosphite triester to a phosphate triester. The coupling efficiency is usually above 99 percent. The yield of full-length product after n couplings with efficiency c is:

For example, with c = 0.99 and a 25-mer (24 couplings), Y ≈ 0.99²⁴ ≈ 0.79, whereas for a 60-mer Y ≈ 0.99⁵⁹ ≈ 0.55. For this reason, longer oligonucleotides are purified by methods such as HPLC or gel electrophoresis. Primers for PCR are typically 18 to 30 nucleotides long.

Choice of DNA

Synthetic primers are almost always DNA. The 2′-hydroxyl group of ribose makes RNA more susceptible to base-catalyzed hydrolysis and to ribonucleases, which are ubiquitous. DNA lacks this group and is stable to the temperatures of PCR (up to about 95 to 98 °C), although it can undergo depurination at low pH and high temperature over long times. Most thermostable DNA polymerases extend DNA primers efficiently, and RNA primers are used only in special cases, such as some isothermal methods and in studies of priming.

Melting temperature and annealing

The primer anneals to its target at a temperature that is related to its melting temperature T_m, at which half of the primer molecules are in the duplex with the target. For a short oligonucleotide, T_m can be estimated by simple rules, such as the Wallace rule for primers of 14 nucleotides or less:

where N_X is the number of nucleotides of each type. This is only a rough guide. Modern programs use the nearest-neighbor model, in which the free energy of the duplex is the sum of the contributions of each adjacent pair of bases, plus corrections for initiation and the ends:

The enthalpy and entropy are summed in the same way, and the melting temperature for a primer at a total strand concentration C_T is then (for non-self-complementary strands):

where R is the gas constant, _ΔH_° is the total enthalpy of formation, and _ΔS_° is the total entropy. The melting temperature increases with the concentration of strands and with the concentration of monovalent and divalent cations, which reduce the electrostatic repulsion of the phosphate backbones. It is decreased by additives such as dimethyl sulfoxide and formamide. The annealing temperature of the reaction is usually set a few degrees below the lower of the two primer T_m values.

Design rules

Common practical criteria for PCR primers include:

• A length of about 18 to 25 nucleotides, and a T_m of about 55 to 65 °C, with the two primers matched to within a few degrees.
• A GC content of about 40 to 60 percent, and a "GC clamp" of one or two G or C bases at the 3′ end to promote stable binding, but not more than three or so at the end, which can cause mispriming.
• No long runs of a single base, and no repeats of dinucleotides.
• Specificity for the target, checked by sequence alignment of the primer against the genome (for example with BLAST or primer design software).
• No strong secondary structure in the primer or in the template region to which it binds.

Two kinds of unwanted structure are important. A primer with a self-complementary region can fold into a hairpin, which has a stable stem if the free energy of folding is sufficiently negative. A primer with a region that is complementary to another primer, or to a copy of itself, forms a primer dimer. If the dimer has complementary 3′ ends, the polymerase extends it, which makes a short product that competes for polymerase, nucleotides, and primers and reduces the yield of the desired product. Hot-start polymerases, which are inactive until the first high-temperature step, reduce this problem.

Degenerate and modified primers

Primers can be modified in many ways because they are made by chemical synthesis:

• Degenerate primers contain mixtures of nucleotides at some positions (using IUPAC codes such as R for A or G). They are used to amplify related sequences, for example when a gene is known only from the protein sequence of a homolog. Inosine, which pairs with several bases, is also used.
• Reporter modifications. Fluorescent dyes, biotin, and other labels are attached at the 5′ end, where they do not block extension. Labeled primers are used in fragment analysis, sequencing, and capture methods.
• Tails. A 5′ extension that does not anneal to the target can carry restriction enzyme recognition sites, recombination sequences, promoter sequences, or adapters. The tail is copied into the product, which allows cloning and library preparation. Because the 3′ part determines annealing, the 5′ tail does not affect specificity.
• 3′ modifications. A blocked 3′ end (for example a 3′-phosphate or a dideoxynucleotide) prevents extension, and is used in blocking primers and in some allele-specific methods.
• Backbone and sugar modifications. Phosphorothioate linkages resist nucleases, and locked nucleic acids (LNA) increase the T_m and the discrimination of mismatches.

Primers in Other Methods

• Sequencing. In Sanger sequencing, a single primer is extended in the presence of chain-terminating dideoxynucleotides. In sequencing by synthesis, a primer anneals to an adapter sequence.
• Quantitative PCR. A pair of primers is used, sometimes with a labeled probe or an intercalating dye. Reverse transcription of RNA is primed with random hexamers, with oligo(dT), which binds the poly(A) tail of mRNA, or with gene-specific primers.
• Isothermal amplification. Loop-mediated isothermal amplification (LAMP) uses four to six primers, and rolling circle amplification and recombinase polymerase amplification use their own priming strategies.
• Multiplex PCR. Many primer pairs are used together, which requires that all have similar T_m values and do not interact.
• Primer extension assays. A labeled primer is extended on an RNA or DNA template to map the 5′ end of a transcript or the sites of modification.

Why does a single nucleotide mismatch at the 3' end of a primer block PCR amplification far more severely than an internal mismatch?

Replicative and thermostable DNA polymerases inspect the minor-groove geometry of the terminal base pair in the active site before catalyzing nucleophilic attack by the 3'-OH group. While an internal mismatch slightly lowers the overall melting temperature (T_m) of the duplex without altering the 3' terminus, a 3'-terminal mismatch distorts the catalytic alignment of the 3'-OH relative to the incoming dNTP α-phosphate and Metal A, reducing extension rates by 10²- to 10⁶-fold.

Why does solid-phase phosphoramidite chemical synthesis build primers in the 3'-to-5' direction rather than the biological 5'-to-3' direction?

In solid-phase phosphoramidite chemistry, the growing oligonucleotide is anchored to controlled-pore glass (CPG) at its 3' terminus while the reactive phosphoramidite group is attached to the 3' carbon of the incoming monomer and a acid-labile dimethoxytrityl (DMT) protecting group masks its 5'-OH. Because a primary 5'-hydroxyl group is sterically less hindered and far more nucleophilic than a secondary 3'-hydroxyl group, 3'-to-5' coupling achieves >99% per-step efficiency.

How does the end-replication problem arise from RNA primer removal on linear chromosomes?

On the leading strand of a linear eukaryotic chromosome, DNA polymerase ε can synthesize DNA continuously to the very tip of the template. On the lagging strand, however, once RNase H2 and FEN1 excise the final 5'-terminal RNA primer at the chromosome end, no upstream 3'-OH group exists to prime gap-filling DNA synthesis, leaving a 3' single-stranded overhang and shortening the chromosome by 50 to 200 base pairs per cell division unless extended by telomerase.

• DNA replication
• DNA polymerase
• Primase
• Polymerase chain reaction
• Oligonucleotide synthesis
• Okazaki fragment
• Reverse transcriptase