How we learned what genes are made of

In the 1940s, scientists made a discovery now fundamental to biology: genes are encoded in DNA. The story involves bacteria, dead mice, and a kitchen cream separator.

This is the first of three pieces from Works in Progress Issue 24 that will go out ahead of the magazine arriving with subscribers – other articles will come out after the magazine arrives. Subscribe by 1st June and get the print edition when its released in the second week of June.

---

In the TV miniseries Lessons in Chemistry, chemist Elizabeth Zott presents her research on de novo nucleotide synthesis to a panel of suited and bespectacled colleagues. ‘Unlike the amino study group’, says Zott, ‘we are starting with the basic assumption that DNA, not protein, is the basic foundation of life’. The panel scoffs at this apparently ridiculous claim. The head of her department dismisses DNA as a ‘dead end’ and Zott’s method for making it from scratch as ‘nothing more than a party trick’. Zott’s proposal is rejected.

Zott was a fictional character, but the scientific debate was real. The fact that DNA encodes genetic information is now taught in biology classrooms worldwide, but until the late 1940s it was a fringe idea. Most scientists instead believed genes were made of protein. James Watson and Francis Crick are now household names for discovering DNA’s double helix structure, but the importance of that discovery rested on earlier work establishing that DNA, not protein, carries genetic information. The scientists who made this more fundamental discovery have often been overlooked.

The organisms that enabled this discovery were similarly unexpected: bacteria. At the time, few scientists, even microbiologists, thought that bacteria could offer anything of value to genetics. Many doubted they even had genes! Those primitive blobs swimming under microscope lenses were thought to be nothing more than tiny bags of enzymes – a totally different kind of life. So it came as a surprise when the pivotal discovery of modern genetics emerged not from a geneticist but a microbiologist.

What parents pass on

Evolution by natural selection requires that organisms inherit characteristics from their parents. Heredity wasn’t a new idea when Darwin published On the Origin of Species in 1859. Farmers had exploited it for centuries, selectively breeding livestock and crops for desirable traits. But neither Darwin nor anybody else at the time could explain how traits were passed down from parents to offspring.

Inheritance was full of mysterious patterns: skin color tended to be a blend of the parents’, eye color could differ from either, and sex only matched one. How could such variation be explained? There were plenty of theories – including Darwin’s own provisional hypothesis of ‘pangenesis’, in which every cell in the body shed tiny particles called ‘gemmules’ that traveled to the reproductive organs and were passed to offspring – but nobody had persuasive evidence to support one over another.

Nobody, that is, except a little-known Austrian monk named Gregor Mendel. Mendel’s experiments with pea plants showed that inheritance depended on discrete factors transmitted from parents to offspring. Some traits, such as peas being wrinkled rather than smooth, could disappear and then pop up again in later generations – an impossibility if traits were a simple blend of both parents. Mendel’s work went unnoticed in his lifetime, but in 1900 three botanists independently rediscovered and confirmed his findings. By 1909 these factors had been given a name: genes. But the original meaning differed significantly from today’s. Wilhelm Johansen, who coined the term, explicitly rejected the idea that genes were physical particles, instead conceptualizing them as a chemical or physiological process.

The geneticist Thomas Hunt Morgan hypothesized that genes might either be a ‘chemical molecule’ or a ‘fluctuating amount of something’, but concluded: ‘I see at present no way of deciding’. By 1933, the confusion hadn’t resolved. Morgan wrote in his Nobel Prize lecture: ‘There is no consensus of opinion amongst geneticists as to what the genes are – whether they are real or purely fictitious’.

A common thread was the assumption that genetic information must be encoded in protein, not DNA. Proteins are made from long chains of simpler molecules called amino acids, joined in sequence like beads on a string. With 20 different amino acids for each position, like letters in an alphabet, even short protein sequences can produce an astronomical number of combinations. Meanwhile, the alphabet of DNA has only four ‘letters’, made from the four simpler molecules called nucleotides that join in sequence to make strands of DNA. Today, a four-letter alphabet may seem luxurious compared to the binary language of computers (1 and 0), but while information theory would eventually come to influence biologists’ thinking, this would not take hold for a few decades more.

Further, the best evidence suggested that the four letters of DNA were present in equal proportion in every organism, and arranged in repeated blocks. Such a monotonous molecule could be structural, but surely not informational. The physicist-turned-biologist Max Delbrück spoke for many at the time when he derided DNA as ‘so stupid a substance’.

For decades, geneticists remained stuck. To establish causality, they needed to show that a chemical induced predictable and hereditary changes in cells. The dream experiment would be to isolate and purify either DNA or protein from one organism, introduce it into another, and observe a heritable change. But living things could not take up genetic material from their environment – or so they thought. That assumption would be overturned by experiments with bacteria.

An unlikely answer

For much of the early twentieth century the study of bacterial variation and heredity was hopelessly confused. As late as 1942, the biologist Julian Huxley excluded bacteria from his theory uniting Mendelian genetics with Darwinian natural selection on the grounds that ‘they have no genes’. It seemed unlikely that bacteria had anything to offer genetics.

Then, in 1928, the British bacteriologist Frederick Griffith published a paper that shocked researchers around the world. Griffith was studying pneumococci, the bacteria responsible for pneumonia, and a promising new treatment called ‘serum therapy’ in which antibody-rich serum from patients who survived an infection could be used to treat others. But subtle differences between pneumococci rendered serum from one infection ineffective against others, so scientists categorized pneumococci into ‘serotypes’ (I, II, III, and IV). A pneumococcus’s type was fixed and inherited: it was genetic. The bacteria also differed in their colony shapes. Some produced smooth (‘S’) dome-shaped colonies, while others were rough (‘R’) and irregular. The S form was virulent and deadly, thanks to a slippery coat that protected it against our immune cells. But the R form, lacking this coat, rarely caused disease.

Pneumococcus colonies before and after their transformation. In the left image, we see type II-R colonies. Afterwards, they exhibit characteristic type III-S colony morphology. Source: Avery et al. (1944)

In his experiment, Griffith combined two bacterial strains that should have been harmless when mixed: live type I bacteria of the rough form, which didn’t cause disease, and heat-killed type II bacteria of the smooth form, which had once been virulent but were now inert. Injected into mice separately, neither was harmful. But together, the mixture proved unexpectedly deadly. Even more surprising, when Griffith isolated bacteria from the dead mice, he recovered live type II smooth pneumococci. Somehow, by coming into contact with the remains of the virulent dead bacteria, the harmless live bacteria had been ‘transformed’ to match the dead ones. And this change was not temporary: the transformed bacteria continued to produce descendants of the same type as they multiplied. It was heritable.

Immunologists were initially skeptical of Griffith’s work. Perhaps some small fraction of the virulent pneumococci hadn’t really been killed? But soon the experiment was confirmed independently by labs around the world.

Among those struck by Griffith’s experiment was Oswald Avery at The Rockefeller Institute. For 20 years, Avery had studied the immunology of pneumococci. After reading Griffith’s paper, he assigned trainees to investigate transformation. One was JL Alloway, who showed that transformation could also be brought about in vitro with chemical extracts from the killed smooth cells. By dissolving living cells and filtering out cellular fragments, Alloway extracted the ‘thick syrupy precipitate’ that contained whatever it was that was causing bacteria to change serotype. This mixture became known as the ‘transforming principle’.

Solving the puzzle

Avery had shown that the transforming principle caused hereditary changes. That meant it contained genes. If he could isolate the component responsible for transformation and determine its chemical composition, he’d finally have an answer to the long-sought question of what genes were made of.

But on the precipice of a major scientific breakthrough, progress suddenly stopped. Avery was diagnosed with Graves’ disease, a thyroid condition that exhausted him and gave him a tremor that made it difficult to perform experiments. After having his thyroid surgically removed he spent months recovering; when he finally returned to work, he shifted his research focus to newly discovered antibiotics.

When Avery came back to the question of the transforming principle eight years later, the field was largely unchanged. Why no other group swooped in during his absence remains a mystery. Perhaps geneticists were uninterested in working with bacteria, or bacteriologists had prioritized practical research into vaccines and antibiotics.

Whatever the reason, Avery began by refining the method to extract the transforming material with the help of scientists Colin MacLeod and Maclyn MacCarty. Their new process required vast quantities of bacterial culture: 75 liters of broth teeming with bacteria to obtain just 10 to 25 milligrams of transforming principle. They adapted a steam-driven kitchen cream separator to separate bacterial cells from the broth. Obviously, the device was not intended for handling pathogenic bacteria. Tiny gaps in the seals meant that its use filled the room with an invisible mist of potentially lethal pneumococci. Avery’s team had to construct a vessel that could contain and sterilize the machine before opening. The cake of bacteria it collected had to be handled with towels soaked in germicide, and the recovering Avery left the lab whenever it was being used.

Through this effort, the team purified a substance that had rich transforming activity. It contained a mixture of polysaccharides, proteins, RNA, and DNA. Next, they needed to work out which of these was responsible for transformation. Adding enzymes that inactivated proteins or polysaccharides did not stop the bacteria from transforming. But when they added an enzyme that destroyed DNA, transformation no longer occurred. That meant the transforming principle, and therefore the gene, was DNA.

The mood in the lab was electric. But Avery knew it would be an uphill battle to convince others that DNA, not protein, was the genetic material. Opponents would argue that minute traces of protein remained and these were what accounted for the transforming activity. Since it was impossible to prove a negative (that the samples were protein-free), Avery aimed to gather as much evidence as he could: chemical composition, enzyme inactivation, centrifugation, electrophoresis, and ultraviolet absorption. Every result converged on DNA.

In a letter Avery wrote to his brother, he overflowed with excitement about the implications of his discovery. ‘It touches genetics, enzyme chemistry, cell metabolism and carbohydrate synthesis, etc.’ But in his public addresses, he was cautious to a fault. ‘It’s lots of fun to blow bubbles – but it’s wiser to prick them yourself before someone else tries to’.

Avery, MacLeod, and McCarty’s paper was published in 1944. Reflecting Avery’s caution, the word ‘gene’ does not appear in the title, summary, or conclusion, and the paper ends by acknowledging that minute amounts of contaminants nevertheless could be the true source of the transforming activity. But the conclusion was unmistakeable: genes were made of DNA. The response to the paper was mixed: many praised Avery’s work, while some critics argued that contamination by protein could explain the results. But the burden of proof now rested on them to identify such a protein, and nobody was able to do so.

The molecule that changed everything

Avery’s experiment not only settled the question of the physical nature of genes, but also made it possible to ask a hundred more. How does DNA encode genetic information? Do other living organisms also use DNA, or just pneumococci? What is DNA’s molecular structure, and how does it replicate itself?

Hundreds of papers would be published on the nature of DNA over the following decades. Their authors are among the most celebrated scientists of the twentieth century, including the Nobel laureates Joshua Lederberg, who studied how bacteria can exchange DNA; Alfred Hershey and Martha Chase, who demonstrated that DNA, not protein, is transferred from bacteriophages to bacteria during viral infection; and of course Watson and Crick, who discovered the double helix structure of DNA.

Avery however never received a Nobel Prize, despite having been nominated 38 times. The Nobel committee, unaccustomed to Avery’s ‘restraint and self-criticism bordering on the neurotic’, as described by his protégé René Dubos, perhaps decided it best to wait for further confirmation. But Avery was already 65 when he published his DNA work, and died 11 years later; Nobels are not awarded posthumously. Avery’s work has received less notice than that of other geneticists, in large part because he never received science’s top prize. Nevertheless, it is the foundation upon which the last 75 years of genetic and biomedical research has been built.

DNA now dominates all facets of biomedical research. But perhaps the most important consequence of Avery’s investigations has been the recognition that all living things on Earth – from deep-sea algae to high-flying birds, sharks to iridescent beetles, disease-causing bacteria to humans – store their genes in the same way: in the twisting strands of DNA.

---

Kevin Blake is scientific editor at Washington University in St Louis.