The DNA Riddle: King's College, London, 1951-1953

When Franklin arrived at King's College in 1951, J. T. Randall's biophysics program had been in existence for only five years. Like many researchers of that era, Randall, the inventor of radar, was interested in using physics to determine the structure of large molecules so as to understand how they functioned in larger organisms. The molecules involved in heredity were of special interest. Though Oswald T. Avery had proposed that DNA was the key chemical substance of heredity, many scientists believed that proteins were more likely candidates. Randall's program aimed to investigate all aspects of "the physical factors affecting mitosis and cell division, by direct and indirect methods."

Franklin's fellowship proposal called for her to work on x-ray diffraction studies of proteins in solution. However, there was a shift in research priorities after Maurice Wilkins, the assistant director of Randall's lab, began working with an unusually pure sample of DNA obtained from Rudolf Signer. Excited about the possibilities, Wilkins suggested to Randall that Franklin's expertise might be better applied to this promising DNA research. Randall agreed; he wrote to Franklin in November 1950, explaining the change of plan, and stated that she and graduate student Raymond Gosling would be the only staff doing crystallographic studies of DNA. Randall did not mention Wilkins' serious interest in DNA, nor did he tell Wilkins the details of the letter. These omissions soon generated misunderstandings between Wilkins and Franklin--Franklin assumed that the x-ray diffraction studies of DNA would be her project alone; Wilkins assumed that she was joining the loosely organized research team ("Randall's Circus") at the biophysics lab, as the expert on crystallography. When Wilkins continued working on DNA and suggested that he and Franklin collaborate, she resented what she regarded as interference. The situation was exacerbated by differences in personality. Franklin had always been direct, honest, and unafraid of argument; indeed, with her colleagues in Paris, she learned to enjoy "a good row." As friends and family often noted, she could be kind, generous, and fun, but she did not suffer fools gladly. Wilkins had a gentler, more reticent personality, and found Franklin's manner intimidating at times. Besides this, Franklin had been reluctant to return to England and to what she viewed as a stultifying intellectual and social climate. Though Randall's staff included a number of women, King's College, founded as a seminary for Anglican clergymen, was still a largely male preserve, maintaining a separate common room for men, and another which was open to both men and women. Franklin sorely missed the egalitarian atmosphere of Paris, and may have kept her distance from the team in London simply out of homesickness for "le labo." As a result, Franklin and Wilkins did not collaborate to any great extent as the work on DNA went forward.

The material that Franklin was to study was calf thymus DNA, extracted and purified by Rudolf Signer of Berne, and given to several researchers, including Wilkins, at a conference in May 1950. Signer had used an extremely gentle technique that avoided breaking up the large molecules. Preparing some of it for optical studies with a reflecting microscope, Wilkins discovered that this DNA could be drawn out into very thin, uniform fibers like a spider web. This suggested to him a regular structure that might yield an x-ray diffraction pattern, so he and Gosling improvised an apparatus and took photos of the samples. Varying the relative humidity gave them several different results, some showing a disordered pattern and others showing a crystalline pattern. Meanwhile further clues about the possible shape of DNA had come from infra-red spectroscopy studies at King's and the work of Sven Furberg at Birkbeck College (in Furberg's models, the DNA was in helical form, with the sugars perpendicular to the bases and the bases on the inside). In May of 1951, Wilkins gave a paper at a meeting in Naples in which he outlined this and other work being done at King's. (One of those attending the presentation was James Watson, who was intrigued by the findings.) By July, Wilkins was considering that the DNA molecule might be helical, like the polypeptide chain of the protein alpha-helix recently discovered by Linus Pauling. He asked Alec Stokes, a theoretical physicist at King's, to determine mathematically what sort of x-ray diffraction pattern a helix would produce; Stokes rapidly produced a pattern with a strong central "X" much like the one Wilkins and Gosling had made. When Wilkins presented these ideas at a small meeting in Cambridge that month, however, Franklin quietly told him afterwards to leave the diffraction work to her and (as Wilkins later recalled) to go back to his microscopes. Though shaken and confused by this, Wilkins still expected that their relations might improve. In September, he returned from several summer conferences excited about what he had learned, and with a vial of DNA from Erwin Chargaff's lab, which he hoped would help expand the DNA work.

In the meantime Franklin had assembled her new x-ray equipment and Phillips micro-camera and settled down to work. She solved the problem of keeping the humidity constant inside the camera by bubbling hydrogen through salt solutions to the desired humidity levels. Working with single DNA fibers at very high humidity, she discovered that there were two forms of DNA: the familiar "dry" crystalline form ("A") and a longer, thinner, heavily hydrated "paracrystalline" form, which she and Gosling called the "B" or "wet" form. Samples could shift from one form to the other if humidity levels changed. (The existence of two distinct DNA forms explained why earlier attempts at diffraction pictures, such as William Astbury's, had been so fuzzy: the samples contained both forms.) Franklin noted, "Either the structure is a big helix or a smaller helix consisting of several chains. The phosphates [of the sugar that forms the backbone of the molecule] are on the outside so that phosphate-phosphate inter-helical bonds are disrupted by water." The hydrophilic phosphates caused the molecule to soak up water and lengthen. The wet DNA produced a sharp diffraction picture that resembled the pattern Stokes had predicted. Very pleased when he heard of this, Wilkins suggested that he and Stokes collaborate with Franklin. She angrily refused, and their relations became quite strained. At Randall's prompting, they reached a compromise: Franklin would work on the A form, using the Signer DNA, and Wilkins would work on the B form, using the Chargaff DNA. The Chargaff sample, which had been degraded during extraction, turned out to be unsuitable for diffraction studies and Wilkins was unable to make much progress.

Franklin and Gosling continued to photograph and analyze samples of DNA in the A form. The A form presented a different picture, far from clearly helical. Though it was hard to think how DNA could be helical in one form but not in another, the mathematical analysis that Franklin and Gosling carried out did not support a helical structure. Franklin's notebooks from 1951-1952 show that she thought a helix possible, but her data on the A form did not yet confirm it, and she would not theorize in advance of the evidence. To her, the proper approach was to gather data first and then build models from them, not the other way round.

James Watson and Francis Crick at Cambridge University's Cavendish Laboratory had meanwhile been taking the other way round, attempting to build a model of DNA based on what was already known, and on what current research, including that at King's, suggested. They believed that DNA was helical, and in November 1951 built a model of a three-helix molecule with the phosphates on the inside. The biophysics staff at King's, including Franklin, were invited to Cambridge to see it. Franklin immediately noted, correctly, that such a configuration would not hold together. In response, the director of the Cavendish, Lawrence Bragg, ordered Watson and Crick to cease work on DNA and leave it to King's. They did so, going so far as to send their model parts down to the King's staff, though Franklin had little use for them.

During 1952 Franklin took and analyzed ever-sharper photographs of the A form but still didn't see a helix. Her mathematical analysis, the Patterson function, though consistent with a helix in some ways, seemed to show a structural repeat of a dimension that would make a helical folding of the molecule impossible. In July, she and Gosling posted a prank notice of the "death of DNA helix (crystalline)." Franklin felt obliged to consider non-helical structures for the A form. She also believed that the A form, being more crystalline, would yield more precise information. Because DNA could shift between the two forms, any model would need to account for both forms. Yet she at no time argued that the B form was not helical, and this was reflected in the report that was made to the Medical Research Council (which funded the Biophysics Unit) late that year.

Watson and Crick had not stopped thinking about DNA, and they were in regular communication with Wilkins, eager to learn whatever they could of the progress at King's. In January 1953, spurred by Linus Pauling's publication of a 3-helix model (similar to the one they made in 1951), they resumed work on their DNA model, determined to get it right before Pauling or someone else did. Two pieces of evidence from Franklin's work were crucial to their correct model: first, a very clear photo of the B form taken in May 1952 labeled "51" which Gosling had given to Wilkins as part of his graduate research work, and which Wilkins showed to Watson without Franklin's knowledge; and second, the MRC report, given to Watson and Crick by Max Perutz, a member of the MRC committee that reviewed the work at Randall's lab. The report contained details of Franklin's work (as yet unpublished), including her identification of the unit cell as belonging to the crystal space group called face-centered monoclinic C2. The photo confirmed the helical pattern, and the unit cell type told Crick, a physicist with more theoretical crystallography expertise than Franklin, that the helices ran in opposite directions. By early March, they had their model.

Franklin, still unhappy at King's, had arranged to transfer to J. D. Bernal's lab at Birkbeck College, and was hurrying to finish writing up her work on the A form before leaving. She was unaware of the "race for the double helix" that was in process. In February 1953, however, she looked again at photo #51 and began analyzing it. Several days later she concluded that both A and B forms were two-chain helices, although she had not resolved the configuration of the bases inside. She and Gosling drafted an article on the likely molecular structure by mid-March. This appeared, in expanded and modified form, with Watson and Crick's announcement in Nature on April 25, but the draft was done before they had heard about the Watson-Crick model. When Franklin saw the model, she readily accepted it. She and Gosling soon tested the model against their diffraction data for both A and B forms and found that it fit very well. They published an article on these findings in July 1953. Though she subsequently enjoyed very cordial relationships with both Crick and Watson, they never fully acknowledged the important role played by her x-ray diffraction data in their discovery. Shortly after Watson and Crick's announcement, Franklin left the King's lab and DNA work, and moved to Birkbeck College to study virus structure.