Heat, cool, copy, repeat
The claim is a practical one: that a specific stretch of DNA, however small the starting sample, can be copied into enough material to study, by cycling it through a repeating sequence of temperatures rather than growing it in a living cell. The reaction runs in three repeated steps — denaturation, which heats the sample enough to separate the two DNA strands; annealing, which cools it so that short synthetic primers bind to the target sequence; and extension, at an intermediate temperature, where an enzyme builds a new complementary strand from each template. Each full cycle roughly doubles the amount of the targeted sequence, so a typical run of twenty to forty cycles turns a handful of starting molecules into an amount large enough to detect, sequence or clone directly.
A drive up the coast and a lab in Emeryville
Kary Mullis conceived the method in 1983 while working at Cetus Corporation in Emeryville, California, reportedly working out the idea while driving near his home in Mendocino County. He demonstrated the approach successfully before the end of that year, though colleagues at Cetus were initially sceptical, and his supervisor Thomas White reassigned him to the project full-time once the potential became clear. Other scientists at Cetus, including Randall Saiki, Henry Erlich and Norman Arnheim, ran parallel work confirming the method could amplify specific genes from human DNA, publishing a paper on amplifying the beta-globin gene in Science in 1985 that later received its own recognition as a landmark in the field.
The enzyme from a hot spring
What made the technique genuinely practical, rather than merely clever, was a substitution made in 1986: replacing the original DNA polymerase, which the heat of each denaturation step destroyed and which therefore had to be manually replenished after every cycle, with Taq polymerase, an enzyme drawn from Thermus aquaticus, a bacterium that lives naturally in hot springs at temperatures between roughly 50 and 80 degrees Celsius. Because Taq polymerase tolerates the heat that PCR’s denaturation step requires, the whole cycle could run automatically in a single machine without intervention, which is what turned the reaction from a laboratory demonstration into a routine, scalable tool. Mullis and Michael Smith shared the 1993 Nobel Prize in Chemistry for the DNA manipulation work that PCR represented.
What amplification can’t tell you on its own
The method has clear preconditions and failure modes. Designing the short primers that bracket the target sequence requires already knowing enough about that sequence to write them, so PCR cannot amplify a completely unknown stretch of DNA out of nothing. Because the reaction copies whatever template DNA is present with tremendous efficiency, even minute amounts of contaminating DNA can be amplified alongside, or instead of, the intended sample, producing misleading results if laboratory technique is not rigorous. The polymerase used also introduces occasional copying errors, and inhibitory substances present in some real-world samples, such as humic acids in environmental material, can suppress the reaction outright.
One method, a thousand diagnoses
The reach of the technique beyond the original research laboratory is broad and well established. In medicine it underlies prenatal testing, carrier screening for inherited conditions, and detection of cancer-associated mutations; in infectious disease it is the basis for identifying pathogens including HIV, tuberculosis, and, through its RT-PCR variant, the virus responsible for COVID-19. Forensic laboratories use it to generate DNA profiles from crime-scene material too small for older methods, and researchers use it routinely for cloning, sequencing preparation and evolutionary comparisons across species. The shared thread across all of these applications is the same basic cycle described above, adapted rather than replaced for each purpose.
The chemistry earned the prize; the later claims did not
PCR earns its place as a genuine turning point in molecular biology, and the chemistry behind it is not in dispute: the amplification works as described, the enzyme substitution that made it practical is well documented, and its diagnostic and forensic applications are in continuous, verifiable use. What deserves a separate judgement is Mullis himself, whose later public statements questioning the link between HIV and AIDS, and expressing scepticism about human-caused climate change, were not supported by research he conducted and stand apart from the chemistry that won him his prize. Reading about PCR is worth the time on the strength of the technique alone; treating Mullis’s unrelated later claims as carrying the same authority would be a mistake the material itself does not make.
