Paul Berg: A Life in Five Chapters

Chapter 1 · A Brooklyn Science Club

A teacher who let students find things out

1926 – 1959 · Brooklyn · Penn State · Cleveland · Stanford

Paul Berg was born in Brooklyn in June 1926, the son of Jewish immigrants from Russia; his father was a clothing manufacturer.

He credited his career to a schoolteacher at Abraham Lincoln High School, Sophie Wolfe, who ran an after-school science club in which students were not taught results but were set problems and left to work them out. Berg said the important lesson was that science is something you *do*, with an uncertain outcome, rather than a body of facts to be learned. Two other Nobel laureates came out of the same school.

He served in the US Navy at the end of the Second World War, took his degree at Penn State, and his doctorate in biochemistry at Western Reserve. He did postdoctoral work in Copenhagen and then at Washington University in St Louis with Arthur Kornberg, who was working out how DNA is copied.

When Kornberg moved to Stanford to found a biochemistry department in 1959, Berg went with him and stayed for the rest of his career.

His early work was on how the cell translates the genetic code — specifically on the enzymes that attach each amino acid to its transfer RNA, the step that determines which amino acid a codon actually delivers. It is unglamorous, fundamental biochemistry, and it is the ground from which the later work came.

“The tools of genetic engineering came from bacteria, not from us; what we added was the decision to use them.”

— Attributed to Paul Berg, on the origins of recombinant DNA

Why this matters

Berg credited a school science club that set problems rather than teaching results, and the distinction between doing science and learning it runs through his career.

You have the science club and the biochemistry that followed. What would you ask?

Ask Berg

• “What did that after-school science club actually do differently?”
• “What did you learn from Kornberg?”
• “Why does it matter how an amino acid gets onto its tRNA?”
• “Is science something you do or something you learn?”
• “What made you switch from studying genes to moving them?”

Chapter 2 · The Tools Came From Bacteria

Restriction enzymes, ligase, and what made it possible

1968 – 1972 · Stanford

The idea of joining DNA from two organisms is old. What was missing was any way to cut DNA at a predictable place and join it back.

Both tools turned up in bacteria, found by people not looking for them.

*Restriction enzymes* are a bacterial defence against viruses. A bacterium makes an enzyme that recognises a specific short sequence of DNA — say six bases — and cuts the strand wherever that sequence appears, chopping up invading viral DNA; its own DNA is chemically marked so it is spared. Werner Arber worked out why they must exist, Hamilton Smith isolated one, and Daniel Nathans showed they could be used as tools. They shared the Nobel Prize in 1978.

The crucial feature is that many of them cut the two strands at slightly offset positions, leaving short single-stranded overhangs — *sticky ends*. Any two pieces cut with the same enzyme have complementary overhangs and will pair up, whatever organism they came from.

*DNA ligase* is the cell's own repair enzyme, which seals nicks in a DNA backbone. It joins the paired pieces permanently.

So the whole toolkit of genetic engineering was borrowed. Berg was consistently clear about this: the tools came from bacteria and from people studying bacterial immunity for its own sake, and what his laboratory contributed was the decision to use them for this purpose.

In 1972 Berg's laboratory — the work done principally by David Jackson and Janet Mertz — produced a single circular DNA molecule containing DNA from the monkey tumour virus SV40 joined to DNA from a bacterial virus, lambda. Genes from two organisms that had never exchanged genetic material, in one molecule.

Why this matters

Every tool of genetic engineering was found by people studying bacterial defence against viruses for its own sake — the classic case for curiosity-driven research.

You have the borrowed scissors and glue. What is your question?

Ask Berg

• “Why do bacteria have enzymes that cut DNA at specific places?”
• “What are sticky ends and why do they matter?”
• “Why insist that the tools were not yours?”
• “What did Jackson and Mertz actually do at the bench?”
• “Would this have happened without curiosity-driven research?”

Chapter 3 · The Experiment He Did Not Do

A telephone call from Cold Spring Harbor

1972 · Stanford · Cold Spring Harbor

Making the molecule was the achievement. The obvious next experiment was to put it into *Escherichia coli* — a bacterium that lives in the human gut, grows fast, and would copy the molecule indefinitely, giving unlimited quantities of it. That is how the technique becomes useful.

Berg's laboratory was preparing to do it.

Robert Pollack, a tumour-virus biologist at Cold Spring Harbor, heard about the plan and telephoned Berg. His argument was direct: SV40 is a tumour virus. E. coli lives in people, including everyone in the building. If a bacterium carrying tumour-virus genes escaped — and bacteria escape — nobody could say what would happen. Nobody had any idea.

Berg's reaction is the interesting part. He did not dismiss it, and he did not argue that the risk was small. He concluded that he *could not say* what the risk was, that no one could, and that this was itself a reason not to proceed.

He stopped the experiment. Some of his colleagues thought he was being absurd, and the tumour-virus risk in this particular case turned out, in retrospect, to be very small.

But the reasoning generalises, and that is why it lasted. An unquantified risk is not the same as a proven one — and neither of those is a reason to proceed. When you genuinely do not know, the burden falls on the person proposing the action.

It is also worth noting who raised it. The objection came from a colleague on the telephone, not from Berg himself, and Berg always said so.

Why this matters

Berg stopped because nobody could quantify the risk, not because it had been shown to be large — and that reasoning is what made the pause defensible.

You have the call and the experiment abandoned. What would you ask?

Ask Berg

• “What exactly did Pollack say to you?”
• “Why stop when the risk had not been shown to be real?”
• “What did your colleagues think of the decision?”
• “Who should decide whether an experiment is too risky?”
• “Was the risk in that case actually small?”

Chapter 4 · Asilomar

Four days in February 1975

1974 – 1975 · Washington · Pacific Grove, California

In July 1974 Berg and ten colleagues published a letter simultaneously in *Science* and *Nature* asking laboratories worldwide to *voluntarily defer* certain classes of recombinant experiment until the hazards could be assessed. It is usually called the Berg letter or the moratorium letter.

The request was largely honoured. Laboratories across the world stopped work they were free to continue, on the basis of a letter with no legal force.

In February 1975 about 140 people — molecular biologists, plus lawyers, physicians and journalists, the last of these admitted on the condition that they publish nothing until the end — met for four days at the Asilomar conference centre at Pacific Grove, California. Berg chaired.

The argument was fierce. James Watson thought the whole thing an overreaction. Others thought the restrictions inadequate. The lawyers made a point that concentrated minds considerably: if somebody was harmed by an experiment scientists had themselves identified as risky and had proceeded with anyway, personal liability would follow.

What came out was a graded containment framework. Experiments were sorted by estimated risk, with physical containment requirements rising accordingly, and with *biological* containment — using deliberately crippled bacterial strains unable to survive outside a laboratory — as a second barrier. A small class of experiments was prohibited outright.

The US National Institutes of Health turned the framework into formal guidelines in 1976, and much of the world followed.

The criticism is real and has been made ever since: the meeting considered only *physical* hazard. It did not address the ethical questions, the commercial ones, or the eventual use of the technology on humans. And the public — who would bear any consequences — were not represented at all. Scientists set the rules for scientists.

Why this matters

Asilomar is the reference point for whether a scientific community can regulate itself, and the standing criticism is that it excluded everyone who was not a scientist.

You have four days of argument and a set of rules. What is your question?

Ask Berg

• “Why did laboratories honour a letter with no legal force?”
• “What did the lawyers say that changed the room?”
• “Why was the public not represented at Asilomar?”
• “What is biological containment?”
• “Did the moratorium delay useful work?”

Chapter 5 · What Followed

An industry, a Nobel Prize, and a reference point

1976 – 2023 · Stanford · Stockholm

The feared catastrophe never happened. No laboratory-derived recombinant organism has caused an epidemic. Whether that vindicates the caution or shows it was unnecessary is genuinely arguable, and both cases have been made.

What did happen was an industry. Genentech was founded in 1976 by Herbert Boyer and Robert Swanson; human insulin made in genetically modified bacteria was approved in 1982, and human growth hormone followed. Genetically modified crops, gene therapy, monoclonal antibodies, mRNA vaccines and CRISPR gene editing all descend from the demonstration that a gene can be deliberately moved between species.

Berg shared the 1980 Nobel Prize in Chemistry with Walter Gilbert and Frederick Sanger, who had developed methods for reading DNA sequences.

He remained a public figure on questions about the governance of biology. He supported human embryonic stem cell research and campaigned publicly against restrictions on it in the United States. He spoke on gene editing after CRISPR arrived, and was involved in the discussions around germline modification — arguing consistently that the community should confront a question before, not after, somebody acts on it.

He was married to Mildred Levy for over seventy years, and they had one son. He died in February 2023, aged ninety-six.

Asilomar is now invoked every time a technology arrives that its own creators find alarming — in artificial intelligence, in gene drives, in synthetic biology. Sometimes as a model. Sometimes as a warning that a field asked to judge itself will find in its own favour. Berg himself said, late on, that it worked because the risk was narrow and technical, and that it should not be assumed to transfer.

Why this matters

Asilomar is now invoked whenever a field must decide whether to restrain itself, and Berg's own late view was that it worked because the question was narrow.

You have the industry, the prize and the precedent. What would you ask him?

Ask Berg

• “Does the absence of a catastrophe prove the caution was right?”
• “Should Asilomar be a model for artificial intelligence?”
• “Can a field asked to judge itself ever be trusted?”
• “What do you make of gene editing in human embryos?”
• “Which experiment should not be done now?”

What Berg changed

Berg's 1972 molecule is where genetic engineering begins: insulin from bacteria, genetically modified crops, gene therapy and modern biotechnology all descend from the demonstration that a gene can be moved between species. Asilomar remains the reference point whenever a science must decide whether to restrain itself, and it is argued about every time a new technique arrives.

A debate that continues

Asilomar addressed physical containment and left the ethical, commercial and public questions untouched; critics argue it was scientists granting themselves permission, and that the moratorium delayed useful work for a hazard that never materialised.

Keep exploring — ask Berg

• “Who should sit in the room when a science decides its own limits?”
• “What is the cost of pausing a useful technology?”
• “Which of today's experiments most needs an argument first?”

Related lives

• Rosalind Franklin — The Dark Lady of DNA
• Ian Wilmut — Whose Lamb Had No Father
• Frederick Banting — Who Sold the Patent for a Dollar
• J. Robert Oppenheimer — Father of the Atomic Bomb

Related themes

DNA and genetic engineering · Biotechnology · Science and responsibility