by Matthew Cobb
Above all, some excited researchers decided to extend the Avery group’s findings. On 20 January 1945, Joshua Lederberg, a brilliant 19-year-old who had just begun a postgraduate medical degree at Columbia University, sat down to read an article that had been passed to him by a fellow student, Harriett Taylor. The effect was electric. As he wrote in his diary:
I had the evening all to myself, and particularly the excruciating pleasure of reading Avery ’43 [sic] on the desoxyribose nucleic acid responsible for type transformation in Pneumococcus. Terrific and unlimited in its implications … I can see real cause for excitement in this stuff.30
Inspired by Avery, Lederberg decided to turn to the study of transformation in other bacterial systems and soon discovered that bacteria can have a sexual phase, thereby opening the road to the genetic study of these organisms – in 1959 he won the Nobel Prize in Physiology or Medicine for this work. Equally importantly, Erwin Chargaff, a 39-year-old Ukrainian-born biochemist who had fled the Nazis and was now working at Columbia University, became an outspoken champion of Avery’s discovery. Chargaff immediately focused his research on the chemical composition of nucleic acids, and soon began to challenge Levene’s tetranucleotide hypothesis.
In Paris, the deputy director of the Institut Pasteur, 50-year-old André Boivin, was inspired by Avery’s paper to investigate transformation in a completely different bacterium, Escherichia coli. Boivin published his findings in French in November 1945. Like Avery, Boivin found that the transforming agent seemed to be ‘a highly polymerised thymonucleic acid’. Bolder than Avery, Boivin explicitly argued that ‘we should now look to the nucleic acid component of the giant nucleoprotein molecule that forms a gene, rather than to the protein part, to find the inductive properties of the gene.’31 When Avery was shown Boivin’s paper, he beamed and happily announced to his colleagues over lunch that they now had ‘continental support’.32 Everything seemed to be going their way.
* At the time desoxyribonucleic acid was the accepted name; this was subsequently changed to deoxyribonucleic acid. Other names used in the 1940s included thymonucleic acid and desoxypentose nucleic acid. They all referred to the same thing.
* Both men were victims of the war: Griffith was killed in an air raid in 1941, while Neufeld starved to death in Berlin in 1945.
–FOUR–
A SLOW REVOLUTION
As Europe and America emerged from the Second World War, there was a wave of research on the structure and function of nucleic acids, partly propelled by Avery’s work. Between 1944 and 1947, more than 250 scientific papers were published on nucleoproteins and nucleic acids – about the same number as in the new field of antibiotics – and most of them explored the nature and function of nucleic acids rather than proteins.1 Between 1946 and 1948, four international scientific conferences focused on the question – one in Cambridge, England (1946), two at Cold Spring Harbor Laboratory on Long Island (1947 and 1948), and one in Paris (1948). Nucleic acid structure and function was becoming one of the hottest scientific topics of the postwar world.
In July 1946, the Society for Experimental Biology held a symposium in Cambridge on nucleic acids. One of the speakers was William Astbury of the University of Leeds, a pioneer of the use of X-rays to study crystal structures. Astbury had visited Avery’s laboratory in 1937 and knew all about his work on transformation. Within a few months of Avery’s article appearing in 1944, Astbury told a friend that he had been ‘terribly thrilled’ that Avery had identified the transforming substance as DNA; he thought this was ‘one of the most remarkable discoveries of our time’. Astbury wrote: ‘I wish I had a thousand hands and labs with which to get down to the problem of proteins and nucleic acids. Jointly those hold the physicochemical secret of life, and quite apart from the war, we are living in a heroic age, if only more people could see it.’2
At the Cambridge meeting, Astbury showed X-ray images of DNA that indicated very clearly that a DNA fibre contains repeated elements, but he was unable to conclude anything about the sequence of those elements within the fibre. Astbury closed his talk by presenting the first model of DNA structure, explaining:
A test that cannot long be dispensed with in any enquiry into the structure of a complex molecule is that of trying to build an accurate atomic model on the basis of known sizes and interbond angles. Chemical formulae are no more than a convenient shorthand, and it is always revealing, and often startling, to see what a molecule looks like in space.3
His model was a relatively uniform column. It was neither startling nor correct.
At the same meeting, Professor Masson Gulland from Nottingham summarised the work on nucleic acid structure that had been done during the war and questioned the widely held view that DNA and RNA were boring molecules: ‘there is at present no indisputable evidence that any polynucleotide is composed largely, if at all, of uniform, structural tetranucleotides’.4 There was no evidence that the four bases were repeated like beads on a string. As Gulland put it:
there is, to choose perhaps an extreme case as illustration, no reason why four molecules of a given nucleotide should not be adjacent and be succeeded in the chain by, let us say, a group of molecules of another nucleotide.
Gulland was arguing that the sequence of bases along the DNA molecule might vary.
In another talk, Dr M. Stacey of the University of Birmingham discussed Avery’s suggestion that DNA was the ‘transforming principle’. Stacey accepted that DNA played an essential role but argued that it functioned as an enzyme that used tiny amounts of the polysaccharide from rough strains as a template to begin synthesising the new kind of capsule.5 In contrast, Edgar and Ellen Stedman, from the University of Edinburgh, were adamant that nucleic acids were merely a structural part of the chromosomes, and that protein alone could account for heredity:
The material of which the chromosomes is composed must … be capable of accounting in a broad manner for the hereditary functions of chromosomes … The first of these requirements can be satisfied only by one known type of compound, a protein.6
A few weeks earlier, Cold Spring Harbor Laboratory in the US had relaunched its annual Symposium in Quantitative Biology after a hiatus due to the war. The first two meetings were focused on topics that flowed directly from Avery’s work: ‘Heredity and variation in microorganisms’ (1946) and ‘Nucleoproteins and nucleic acids’ (1947). At the 1946 symposium, Avery’s work was cited by one-third of the speakers, and McCarty presented a paper co-authored by Avery and Harriett Taylor in which he boldly extended their view of the role of DNA in transformation to the whole of life, concluding: ‘these results suggest that nucleic acids in general may be endowed with biologically specific properties not hitherto demonstrable.’7 There was kickback from other researchers, such as Seymour Cohen, who argued in another part of the meeting that ‘the data directly relating nucleic acid to specifically inheritable phenomena are very sparse indeed’.8 Despite widespread interest, the discoveries of the Avery group were not met with unanimous approval.
An even clearer critique of Avery’s interpretation appeared shortly afterwards, when Alfred Mirsky published a widely read article that he co-authored with Arthur Pollister. Mirsky and Pollister pointed out that ‘there can be little doubt in the mind of anyone who has prepared nucleic acid that traces of protein probably remain in even the best preparations’ and that ‘as much as 1 or 2 per cent of protein could be present in a preparation of “pure, protein-free” nucleic acid.’9 This protein remnant could easily account for the effects that Avery’s group attributed to DNA. Mirsky focused on the quantitative estimations of protein in the extracts produced by the Avery group, ignoring the varied kinds of data that suggested that DNA was the sole active component in their extracts, such as the fact that protein-digesting enzymes had no effect on the action of the transforming principle. Mirsky’s criticisms undermined confidence in Avery’s claims, especially among those who were not chemists. One of the most prominent people to accept Mirsky’s argument was
Hermann Muller. In the written version of his 1945 Pilgrim Trust lecture to the Royal Society, Muller accepted that Avery’s finding, if true, was ‘revolutionary’, but indicated that he was personally convinced by Mirsky’s suggestion that undetected ‘genetic proteins’, floating free in the medium, caused Avery’s results.10
At the June 1947 Cold Spring Harbor symposium the new abbreviation of DNA began to replace the cumbersome desoxyribonucleic acid.11 At the meeting, which was attended by 150 people, the French DNA convert André Boivin presented the big-picture implications of Avery’s findings. For two years, Boivin had been publishing evidence from Escherichia coli bacteria that supported Avery’s conclusion that the transforming principle was composed of DNA, showing that the activity of the substance could be destroyed by the enzyme DNase but not by an equivalent enzyme that attacked RNA, RNase.12 After summarising the evidence, Boivin presented a vision for the future of the whole of biology, speculating about the possibility of transferring genes in higher organisms much as had been done in bacteria:
each gene can be traced back to a macromolecule of a special desoxyribonucleic acid. … Thus, this amazing fact of the organization of an infinite variety of cellular types and living species is reduced, in the last analysis, to innumerable modifications within the molecular structure of one single fundamental chemical substance, nucleic acid … This is the ‘working hypothesis’ quite logically suggested by our actual knowledge of the remarkable phenomenon of directed mutations.13
In the discussion of Boivin’s talk, Mirsky explained why he was still not convinced. Although he agreed that there was no chemical evidence that all nucleic acids were the same, he emphasised that the only evidence for the genetic role of DNA came from bacteria. This made it difficult for Mirsky to accept what he recognised were the revolutionary implications for the whole of biology. Mirsky’s critique repeated the argument he had published the year before: small amounts of protein could still be present in Avery and Boivin’s extracts. ‘In the present state of knowledge it would be going beyond the experimental facts to assert that the specific agent in transforming bacterial types is a desoxyribonucleic acid’, Mirsky said. In response, Boivin accepted that it was impossible to be absolutely certain about the chemical composition of any extract, but he then neatly shifted the argument, underlining the varied kinds of evidence that he and the Avery group had presented: ‘it seems to us that the burden of proof rests upon those who would postulate the existence of an active protein lodged in an inactive nucleic acid.’14
At the meeting, the chemist Erwin Chargaff also turned the tables on those like Mirsky who argued that the genetic material was a nucleoprotein, pointing out that there was no evidence that nucleoproteins were really present in cells; it was quite possible that these compounds were formed accidentally in the test tube when an extraneous protein found itself bound to the DNA while the two compounds were being isolated. Chargaff went on to praise ‘the epochal experiments by Avery and his associates’ and sketched out a future research programme that, although not as grandiose as Boivin’s, had the great virtue of being feasible:
If, as we may take for granted on the basis of the very convincing work of Avery and his associates, certain bacterial nucleic acids of the desoxypentose type are endowed with a specific biological activity, a quest for the chemical or physical causes of these specificities appears appropriate, though it may remain completely speculative for the time being. … Differences in the proportions or the sequence of the several nucleotides forming the nucleic acid chain also could be responsible for specific effects.15
Despite Mirsky’s opposition, Avery’s advocates were sketching out how an apparently boring DNA molecule could be as varied as a protein, perhaps through differences in the sequence of the bases.
Masson Gulland – a cheery-looking cove with a centre parting, a moustache, full lips and laughter lines around his eyes – gave a similar talk to his presentation at Cambridge the year before, but did not fully embrace the idea that DNA had specificity: ‘It seems possible that both nucleic acid and protein may contribute to the specificity, and not the protein alone as has often been thought’, was as far as he would go.16 In the discussion, Gulland was asked about the possibility that a DNA molecule might be a helix, held together by the presence of evenly spaced hydrogen bonds between the bases. Gulland, who had been studying hydrogen bonds in DNA, called this idea ‘interesting and stimulating’, but he did not investigate this further – he died in a train crash within weeks of returning to the UK.17 Two other participants who had also been at the Cambridge meeting, Edgar and Ellen Stedman, continued their critique of the genetic role of DNA, even claiming that it was a moot point that nucleic acid was an integral part of the chromosomes.18 In a similar vein, Jack Schultz complained that although the work of Jean Brachet and Torbjörn Caspersson suggested that nucleic acids played a role in controlling protein synthesis, the actual evidence was ‘far from convincing’.19 During the 1940s, Brachet and Caspersson had both found that RNA was involved in protein synthesis, although what exactly it was doing was unclear.20
Some scientists were more enthusiastic about the importance of nucleic acids. In an article that appeared in 1948 in a new genetics journal entitled Heredity, Joshua Lederberg, who had been inspired by Avery’s paper to study bacterial genetics and was still only 22 years old, wrote:
The total absence of all other components is not readily established, although none can be detected with available methods. The criticism of Mirsky and Pollister (1946) should be noted, however. The further chemical characterisation of the transforming principle is one of the most urgent problems of present-day biology, since it behaves like a gene which can be transferred by way of the medium from one cell to another.21
In 1948, four years after Avery’s first paper appeared, Boivin was arguing that all genes are made of DNA, Chargaff had hypothesised that nucleic acid specificity might involve differences in the sequence of bases, and Lederberg was urging his colleagues that characterising the transforming principle was a central task of biology. As Astbury had put it in 1945, this was indeed a ‘heroic age’.
*
Oswald Avery played no part in any of this. He left Rockefeller at the end of 1948 and moved to Nashville to live with his brother, Roy. He received no further official recognition of his part in the discovery of the genetic role of DNA; although he was nominated for a Nobel Prize, the committee apparently ‘found it desirable to wait until more became known about the mechanism involved’.22 This view may have been reinforced by the fact that the Swedish scientific community was not closely following the debate in the US and the UK about the nature of the hereditary material.23 The brief obituary that appeared in the New York Times when Avery died in 1955 did not even mention DNA.24 As science journalists began to report on the growing wave of excitement within the scientific community about DNA, history was rewritten almost as soon as it happened. In January 1949, the New York Times informed the world that a Rockefeller Institute researcher had discovered that ‘genes consist in part of a substance called desoxyribonucleic acid’. The researcher’s name was Dr Alfred Mirsky.25
Research into the transforming principle was continued by Rollin Hotchkiss at Rockefeller and by Harriett Taylor, who had moved to Paris, where she married the geneticist Boris Ephrussi. Taylor extended the number of bacterial characteristics that could be transformed, thereby showing the generality of transformation and its similarities to genetic factors in higher organisms, while Hotchkiss responded to Mirsky’s carping criticism about the protein content of the extracts by carrying out very precise experiments.26
The first public presentation of Hotchkiss’s new data took place in Paris at the end of June 1948. André Lwoff, a bacteriologist at the Institut Pasteur, had organised a small meeting with the rather pompous title ‘Biological units endowed with genetic continuity’ – these ‘units’ were bacteria and viruses. Hotchkiss described results from a range of techniques that were intended to e
liminate proteins from the transforming principle. After treatment, at most 0.2 per cent of the extract was protein; this was within the margin of error of a result of 0.0 per cent, so it was quite possible that there was no protein at all in his samples.27 It was impossible to be more precise. Despite this clarity, when Lwoff summed up the week’s discussions, he insisted that nucleic acids ‘could and should be normally combined with another constituent, most likely a protein.’28 The old ways of thinking died hard.
Although Hotchkiss’s paper was initially published only in French, news of his findings soon began circulating in the US. Another participant at the Paris meeting was the US virus biologist Max Delbrück, and in the spring of 1949 Al Hershey, a member of Delbrück’s informal phage group, presented Hotchkiss’s data as part of the DNA transformation story at a round-table discussion on nucleic acids organised by the Society of American Microbiologists.29 By the end of the 1940s, it was widely known that the levels of protein in the transforming principle were effectively zero.
Boivin, who by now had moved to Strasbourg, began his talk at the Paris meeting by stating that although DNA specificity had been shown only in bacteria, the conclusion was overwhelming: ‘these facts lead us to accept – until formal proof to the contrary – that this specificity is an example of a general phenomenon, which everywhere plays a major role in the biochemistry of heredity.’30 In his conclusion, Boivin reported that in a wide range of organisms – including many animals – the nuclei of different cells in the same organism had the same amount of DNA; similarly, members of a given species all had the same amounts of DNA, whereas eggs and sperm had only half the amount found in normal cells. This was a decisive discovery. It had been known since Sutton’s observations in 1902 that when most sexual species reproduced, the creation of the egg and sperm cells involved the halving of the usual double or diploid complement of chromosomes, so that one set of chromosomes went into each egg or sperm, to form what is called a haploid cell. When the egg and sperm fused to form the new organism, these two haploid components formed a new diploid set of chromosomes. Boivin’s observation that the amount of DNA corresponded to the chromosome complement at different phases was what would be expected of a gene – nothing like this had been found for any protein. In bacteria, plants and animals, Boivin argued, ‘each gene can, in the final analysis, be considered as a macromolecule of DNA.’31
