The Polymerase Chain Reaction
Imagine needing a trillion copies of one exact DNA sequence. Suppose you start from a single hair root or a droplet of blood. Before 1985, this was essentially impossible without cloning DNA in living bacteria. That approach was slow, laborious, and limited. The Polymerase Chain Reaction (PCR) solved this problem with a test tube, a heat source, and a handful of reagents. It turned DNA amplification into an overnight bench procedure. Any laboratory could perform it. PCR matters because nearly every downstream tool in molecular genetics depends on having enough DNA to work with. DNA Structure tells us the molecule is a stable, information-bearing double helix. DNA Replication tells us how cells copy it with exquisite fidelity. PCR borrows the cell's own replication chemistry — a DNA polymerase, primers, and nucleotides. It repurposes this chemistry outside the cell. Temperature cycling drives repeated rounds of copying, instead of cellular machinery. The result is exponential amplification: one molecule becomes billions within hours. The big picture is that PCR converts DNA analysis from a resource-limited, cloning-dependent enterprise into something else. It becomes a fast, targeted, in-vitro reaction. It underlies forensic identification, prenatal and infectious-disease diagnostics, and ancient-DNA studies. It also underlies nearly all modern genome engineering workflows. This includes the sequencing and editing technologies that build directly on it. This article works through exactly how PCR achieves exponential amplification. It covers the thermal cycle that drives it. It also covers the landmark experiments and innovations that established and refined the method. Two worked quantitative examples are included, along with common conceptual traps. Finally, it introduces the scientist who invented it, and surveys its wide modern applications.
- PCR exponentially amplifies a specific DNA segment using repeated cycles of denaturation, annealing, and extension.
- Two short synthetic primers flank the target sequence and define exactly what gets copied.
- A heat-stable DNA polymerase (Taq, from Thermus aquaticus) survives repeated boiling, enabling automation.
- Each cycle in principle doubles the amount of target DNA; after n cycles, yield scales as roughly 2^n.
- Kary Mullis conceived the method in 1983; it was published and commercialized by 1985–1988.
- PCR underlies genetic testing, forensics, infectious-disease diagnostics, and sets up Sanger Dideoxy Sequencing and CRISPR-Cas9 Genome Editing.
PCR exploits the natural chemistry of DNA replication — primers, polymerase, and complementary base pairing — but drives it with temperature cycles instead of cells, so a chosen DNA sequence doubles almost every cycle until it dominates the tube.
The core idea
The Polymerase Chain Reaction is an in-vitro technique for producing many copies of a specific, defined segment of DNA from a complex starting sample. It requires knowing (or guessing) short stretches of sequence flanking the region of interest, because these flanking sequences are used to design two short single-stranded DNA oligonucleotides called primers. Each primer is complementary to one strand of the template DNA and binds — anneals — at a specific location, oriented so that primer extension by a DNA polymerase proceeds inward, across the target region, toward the other primer. The reaction is driven by cycling the temperature of the sample through three phases: denaturation (heat separates the double helix into single strands), annealing (cooling allows primers to base-pair with complementary template sequences), and extension (a thermostable DNA polymerase synthesizes new DNA 5'→3' from each primer, using the single strand as template). Because each newly made strand can itself serve as a template in the next cycle, the amount of target DNA approximately doubles every cycle — this is what makes the amplification exponential rather than linear. After roughly 25–35 cycles, a single starting molecule can theoretically yield over a billion copies. Crucially, only the DNA sequence between and including the two primer binding sites accumulates exponentially; flanking sequence outside the primers is copied only linearly for the first cycles and quickly becomes a negligible fraction of the product. This selectivity is what allows PCR to pull one specific gene, allele, or short genomic region out of an entire genome's worth of DNA and amplify only that target, essentially unlimited times, without needing to clone anything into a living organism.
Mechanism
Step 1 — Denaturation. The reaction mixture (template DNA, two primers, all four deoxynucleotide triphosphates or dNTPs, buffer with Mg²⁺, and a thermostable DNA polymerase) is heated to about 94–96 °C. Hydrogen bonds between base pairs break, and the double helix separates into two single strands. This mimics the strand separation performed by helicase in vivo, but here it is achieved purely thermally.
Step 2 — Annealing. The temperature drops to about 50–65 °C, chosen depending on primer sequence composition. At this temperature, the short primers (typically 18–25 nucleotides) find and hybridize to their complementary sequences on the single-stranded template via Watson–Crick base pairing. The annealing temperature is a critical parameter: too high and primers fail to bind; too low and primers bind nonspecifically to similar-but-not-identical sequences, generating spurious products.
Step 3 — Extension. The temperature rises to the polymerase's optimum, typically 72 °C for Taq polymerase. Starting at the 3' end of each annealed primer, the polymerase adds dNTPs complementary to the template strand, extending in the 5'→3' direction — exactly the same chemistry used in cellular DNA Replication. Extension proceeds until the polymerase falls off or reaches the end of the template.
Cycle repetition and exponential growth. After the first cycle, each original template strand yields one new complementary strand, extending past the far primer site — the products are longer than the final target. From the second cycle onward, some newly made strands are themselves bounded exactly by both primer sites (because they were synthesized using a first-cycle product as template, and extension runs off the end of that shorter product). By the third cycle, these precisely primer-bounded fragments begin to dominate, and from this point their number doubles almost exactly every cycle, since both original and newly made strands act as templates. If N₀ is the number of starting target molecules and each cycle has doubling efficiency E (ideally close to 1, meaning 100% doubling), the number of target molecules after n cycles is approximately:
With perfect doubling (E = 1), this simplifies to N₀ × 2ⁿ. In practice E is slightly less than 1 due to reagent depletion, polymerase fatigue, and reannealing competition in later cycles, which is why amplification plateaus after roughly 30–40 cycles regardless of starting amount — the "plateau effect."
Primer design constraints. Primers must be unique enough in the genome to avoid binding multiple sites, similar enough in melting temperature (Tm) to each other to share one annealing temperature, and free of self-complementarity that would cause primer-dimers. Typical primer lengths of 18–25 bases give a statistically unique address even in a genome of billions of base pairs, since 4^18 (roughly 7 × 10^10) vastly exceeds the human genome size of about 3 × 10^9 bp.
Taq polymerase and thermostability. Early PCR used the Klenow fragment of E. coli DNA polymerase I, which denatures at each 94 °C step and had to be manually replenished every cycle — laborious and expensive. The breakthrough came from Thermus aquaticus, a bacterium isolated from hot springs in Yellowstone National Park, whose DNA polymerase (Taq polymerase) remains active after repeated exposure to near-boiling temperatures. This single substitution made automated thermal cycling possible, since the same polymerase molecules could survive dozens of denaturation steps without being replaced.
Instrumentation. A modern thermal cycler automates this cycling precisely: a metal block holding the reaction tubes is heated and cooled electronically according to a programmed profile, typically completing 30 cycles in 1.5–3 hours. Each cycle usually lasts 1–3 minutes total across its three temperature steps, though extension time scales with the length of the amplified product (roughly 1 kb per minute for standard Taq polymerase).
The evidence
Kary Mullis's original conception and demonstration (1983–1985). Working at Cetus Corporation, Kary Mullis proposed that a specific DNA segment could be amplified by repeated rounds of denaturation, primer annealing, and polymerase extension, using two primers oriented to converge on the target. He and colleagues first demonstrated this experimentally on the human beta-globin gene and the sickle-cell allele, published in Science in 1985 (Saiki et al., "Enzymatic Amplification of β-globin Genomic Sequences and Restriction Site Analysis for Diagnosis of Sickle Cell Anemia"). In that paper, PCR amplified a segment of the beta-globin gene from genomic DNA, after which a restriction enzyme (DdeI) was used to distinguish the normal allele from the sickle-cell mutant allele based on a restriction site the mutation destroys. This was the first published demonstration that PCR could amplify a specific human genomic sequence to a level detectable by simple gel electrophoresis and enable direct genetic diagnosis, without cloning. In this original protocol, the Klenow fragment of E. coli DNA polymerase I was used, meaning fresh enzyme had to be added after every single denaturation step because the enzyme was destroyed by the 94 °C heat. This made early PCR slow, labor-intensive, and prone to error, since technicians pipetted enzyme by hand dozens of times per reaction.
The Taq polymerase breakthrough (Saiki et al., 1988, Science). Randall Saiki, Henry Erlich, and colleagues at Cetus, isolated and purified a thermostable DNA polymerase from Thermus aquaticus — a bacterium originally discovered by Thomas Brock in the hot springs of Yellowstone National Park in the 1960s, tolerant of temperatures near 70–80 °C. Substituting Taq polymerase for Klenow fragment allowed the entire cycling process to be automated in a single closed tube, since the enzyme survived repeated 94–95 °C denaturation steps without degrading. This paper demonstrated dramatically improved yield, specificity, and sensitivity, and enabled amplification of longer target sequences with higher fidelity. It converted PCR from a boutique manual technique into an automatable, robust, general laboratory method — the version of PCR used essentially unchanged in principle today.
Specificity controls and independent validation. Subsequent work across many laboratories confirmed the central prediction of the PCR model: amplification is strictly dependent on the presence of both correctly oriented, complementary primer sites, and yield increases geometrically with cycle number until reagent depletion causes plateau. Control reactions lacking one primer, using primers with mismatched sequences, or omitting template DNA reliably produced no detectable product, confirming that amplification requires specific primer-directed priming rather than nonspecific DNA synthesis. Gel electrophoresis of PCR products from reactions stopped at different cycle numbers showed the expected pattern: a barely visible band after roughly 20–25 cycles, and a strong, saturating band by 30–35 cycles, consistent with exponential rather than linear accumulation.
Impact recognized by the Nobel Prize. Mullis received the Nobel Prize in Chemistry in 1993 for the invention of PCR, an unusually rapid recognition — under a decade after publication — reflecting how quickly the method transformed molecular biology, genetics, forensics, and medicine.
Worked examples
Example 1 — quantifying exponential yield with imperfect efficiency. Suppose a reaction starts with N₀ = 5 copies of a target sequence (a very small forensic sample), and the empirically measured per-cycle efficiency is E = 0.85 (85% doubling rather than a perfect 100%, typical for real reactions due to imperfect primer annealing). After n = 30 cycles:
Compute (1.85)^30 stepwise: log₁₀(1.85) ≈ 0.2672, so 30 × 0.2672 ≈ 8.016, giving (1.85)^30 ≈ 10^8.016 ≈ 1.04 × 10^8.
Compare this to the idealized case of perfect doubling (E = 1): N₃₀ = 5 × 2^30 ≈ 5 × 1.07 × 10^9 ≈ 5.4 × 10^9. The realistic 85%-efficiency reaction yields roughly 10-fold less product than the idealized calculation — illustrating why real PCR reactions are typically run for more cycles (30–40) than the bare minimum, to compensate for efficiency losses, while also explaining why comparing "cycle number" across different primer pairs or templates without controlling efficiency is scientifically unreliable — a key reason quantitative PCR (qPCR) uses standard curves rather than raw cycle counts to estimate absolute starting quantities.
Example 2 — diagnostic interpretation: distinguishing homozygous and heterozygous genotypes by allele-specific PCR band sizes. A genetic counselor is testing a couple for carrier status of a recessive disease caused by a 4-bp deletion in a particular gene. PCR primers flank the deletion site such that the wild-type allele yields a 210 bp product and the deletion allele yields a 206 bp product (4 bp shorter). Gel electrophoresis is run for the mother, father, and their child. Results (band sizes observed per lane):
| Individual | Bands observed | Genotype inferred |
|---|---|---|
| Mother | 210 bp only | homozygous wild-type (unaffected, non-carrier) |
| Father | 210 bp + 206 bp | heterozygous carrier |
| Child | 210 bp + 206 bp | heterozygous carrier |
Reasoning: because PCR amplifies both alleles present in a diploid genomic DNA sample independently and simultaneously (each allele serves as its own template), a heterozygote produces two distinct amplicon sizes in the same reaction, visible as two separate bands after electrophoresis. Since the mother shows only the 210 bp band, she contributed only a wild-type allele to the child; the child's 206 bp band must therefore have come from the heterozygous father. This lets the counselor confirm the child is an unaffected carrier (not a homozygous mutant), and importantly the mother's genotype rules out her having contributed the disease allele — a determination made directly from PCR product sizes without ever needing to clone or sequence the gene, illustrating how amplicon length itself can serve as a genotyping readout (as in modern STR-based forensic and paternity testing).
Common misconceptions & traps
| Misconception | Why it's wrong |
|---|---|
| "PCR creates new DNA sequence / information." | PCR only copies an existing template; it cannot amplify a sequence that is not already present, and it introduces no new genetic information beyond rare replication errors. |
| "Doubling continues indefinitely with more cycles." | Amplification plateaus after ~30–40 cycles as primers, dNTPs, and polymerase activity become limiting — the "plateau effect," not true exponential growth forever. |
| "Any DNA polymerase works the same in PCR." | Ordinary polymerases denature and lose activity at ~94 °C; only thermostable enzymes like Taq (from Thermus aquaticus) survive repeated cycling, which is precisely why Taq's discovery was a milestone, not a footnote. |
| "PCR requires knowing the entire target sequence." | Only short flanking sequences (for primer design) need to be known; the region between primers can be entirely unknown and is exactly what gets discovered/sequenced afterward. |
| "More starting template always means more final product." | Because of the plateau effect, reactions with very different starting amounts can end up with similar final yields — this is why quantitative PCR tracks amplification during the exponential phase, not the endpoint. |
The scientist behind it
Kary Banks Mullis (1944–2019) was a biochemist working at Cetus Corporation in Emeryville, California, in the early 1980s, at a time when synthesizing oligonucleotide primers and running thermal cycling by hand was already technically feasible but had never been assembled into an amplification method. Mullis later described the idea occurring to him during a nighttime drive along the California coast in 1983, recognizing that repeated cycles of primer annealing and extension could double a target DNA sequence exponentially if the strands were separated by heat between rounds. He was, by his own account and by colleagues' recollections, an unconventional and sometimes controversial figure, but the core chemical insight — using two opposing primers and cyclic heating to drive geometric amplification — was rapidly validated experimentally by his Cetus colleagues, including Randall Saiki, Stephen Scharf, and Henry Erlich, who carried out the foundational beta-globin experiments and later the Taq polymerase optimization.
Origins · 1 minKary Banks Mullis and how this was discovered
Where this matters today
Clinical diagnostics. PCR-based tests detect pathogens directly from patient samples by amplifying species-specific DNA or RNA (via reverse-transcription PCR, RT-PCR) sequences, including HIV viral load monitoring, tuberculosis detection, and — most visibly to the public — SARS-CoV-2 diagnosis during the COVID-19 pandemic, where RT-PCR became the reference-standard test worldwide.
Genetic disease screening. Prenatal testing, carrier screening, and newborn screening for conditions such as cystic fibrosis and sickle cell disease use PCR to amplify specific gene regions before mutation analysis, following essentially the same logic as the original 1985 sickle-cell demonstration.
Forensic science. PCR amplifies short tandem repeat (STR) loci from trace biological evidence — a single hair follicle, a drop of blood, saliva on a cigarette butt — generating enough DNA for profiling even from samples far too small for older techniques, underpinning modern criminal forensics and paternity testing.
Ancient DNA and evolutionary biology. PCR allows amplification of minute, degraded DNA fragments recovered from fossils, museum specimens, and archaeological remains, enabling studies of Neanderthal genomes and species phylogenies from material that would otherwise be unusable.
Foundation for downstream technologies. PCR-amplified DNA is the standard starting material for Sanger Dideoxy Sequencing, which requires abundant template to read sequence directly, and for CRISPR-Cas9 Genome Editing workflows, where PCR is used to genotype edited cells, verify guide RNA target sites, and prepare DNA fragments for cloning into editing constructs. Real-time quantitative PCR (qPCR) extends the basic method to measure gene expression levels and viral loads precisely, by monitoring amplification during, rather than after, the exponential phase.
Agriculture and conservation. PCR-based genotyping verifies crop varieties, detects genetically modified organism (GMO) content, and identifies endangered species from environmental DNA (eDNA) in water or soil samples, supporting biodiversity monitoring without capturing organisms directly.
Ask the tutor: How does real-time quantitative PCR (qPCR) use fluorescence during the exponential phase to measure the original amount of starting DNA?
Why does PCR need two primers instead of one?
Two primers are required because DNA polymerase can only extend a strand, never initiate synthesis from scratch; each primer provides the free 3'-OH group the polymerase needs to begin. The two primers are designed to anneal to opposite strands of the double helix, oriented so their extensions point toward each other, converging on the target region. This arrangement is what makes amplification exponential and bounded: each strand made in one cycle can serve as template for the opposite primer in the next cycle, and only the region between (and including) both primer sites gets doubled every cycle, giving PCR its defining specificity.
What would happen if you used a non-thermostable polymerase, like the Klenow fragment, for PCR?
This is exactly what Mullis's original 1985 protocol did, and it worked, but with major drawbacks. Because Klenow fragment denatures irreversibly at the ~94 °C denaturation temperature, fresh enzyme had to be manually pipetted into the reaction after every single cycle, making the process slow, expensive, and error-prone over 20–30 cycles. The 1988 introduction of thermostable Taq polymerase from Thermus aquaticus eliminated this requirement entirely, since Taq survives dozens of heating cycles without losing activity, enabling fully automated, closed-tube thermal cycling — the version of PCR used today.
Does PCR amplify RNA directly?
No, PCR amplifies only DNA templates, because DNA polymerases cannot use RNA as a template. To amplify RNA (for example, viral RNA genomes like SARS-CoV-2, or messenger RNA for gene expression studies), the RNA must first be converted into complementary DNA (cDNA) using the enzyme reverse transcriptase, in a step called reverse transcription. The resulting cDNA is then used as the template for standard PCR; this combined workflow is called RT-PCR, and it is the basis of most RNA virus diagnostic testing.
How is PCR different from DNA cloning in bacteria for producing many copies of a gene?
Bacterial cloning inserts a DNA fragment into a plasmid vector, transforms it into living bacterial cells, and relies on the cells' own replication machinery and growth over generations (often days) to amplify the DNA, yielding a stable, renewable, but slower and more labor-intensive source. PCR instead uses purified enzymes and primers in a test tube, requires no living cells, and produces usable amounts of amplified DNA in a few hours. Cloning is still valuable for long-term DNA storage, expression of proteins, and applications needing genuinely novel recombinant constructs, while PCR excels at rapid, targeted amplification of a known or partially known sequence.
Why is primer design so critical to successful PCR?
Primers determine both the specificity and the efficiency of the entire reaction. If a primer sequence matches multiple sites in the genome, it produces multiple unintended products; if the two primers have very different melting temperatures, no single annealing temperature will work well for both; and if a primer folds on itself or binds its partner primer (forming primer-dimers), it wastes polymerase activity on non-target amplification. Careful primer design — typically 18–25 nucleotides long, with balanced GC content, similar melting temperatures, and no strong self-complementarity — is therefore one of the most important practical determinants of whether a PCR reaction succeeds.
Can PCR errors introduce mutations into the amplified DNA?
Yes, though at a low rate. Taq polymerase lacks 3'→5' proofreading exonuclease activity (unlike many replicative polymerases in vivo), so it makes base-incorporation errors at a rate of roughly one error per 10,000–20,000 nucleotides copied per cycle. Over many cycles and especially over long amplicons, these errors accumulate and can matter significantly for applications requiring very high fidelity, such as cloning a gene for functional studies. High-fidelity engineered polymerases with proofreading activity (derived from other thermophilic organisms) are used when error rates need to be minimized, at some cost to amplification speed.
How does PCR relate to techniques taught right after it, like Sanger sequencing or CRISPR?
PCR is typically the first step in generating enough pure, targeted DNA for these downstream methods. Sanger Dideoxy Sequencing requires abundant single-target DNA as starting material, which PCR readily supplies from even trace samples. CRISPR-Cas9 Genome Editing workflows use PCR to amplify the genomic region around a target site both before editing (to design guide RNAs and confirm sequence) and after editing (to verify that the intended cut and repair occurred correctly). In this sense PCR is less a single application and more a foundational amplification step that many later molecular genetics techniques depend on.