Rosy Revised

The molecular revolution in biology began 50 years ago with the discovery of the structure of DNA, and has had such an impact that the reading public’s interest now extends even to the lives of molecular biologists. Indeed, molecular biologists have stolen the limelight from physicists and astronomers. Best known among them are the Nobel laureates James Watson and Francis Crick; less well known is Rosalind Franklin, who died in 1958 aged 37. Today many believe that, had she lived, she, too, would have won a Nobel Prize for her pivotal contribution to the work on DNA and subsequently on the structure of the tobacco mosaic virus. In 2000, as if to mark a new beginning, King’s College London dedicated a new building in the Strand to Franklin and Maurice Wilkins, the third recipient of the Nobel Prize, who carried out their DNA research in the College’s biophysics unit. Brenda Maddox calls this ‘a genuflection bordering on political correctness’.

In The Double Helix (1968) James Watson makes clear his personal view of how science should best be done: in a spirit of rivalry and competition, and not one of consensus as to the allocation of effort and resources. Watson knew that the manner in which he and Crick had arrived at their proposed structure, the ‘double helix’, would make a great story. It had both suspense and secrecy, and involved the semi-covert use of Franklin’s data. Watson seemed oblivious to the ethical aspects of the story he told so candidly. His style was crisp, breezy and at times unfair to those he described – especially to Rosalind Franklin. His Franklin was a bluestocking, lacking the ‘judgment’ that would have led her to the theory needed to interpret her data. She was ‘sticky’, she had an ‘acid smile’, her voice lacked ‘warmth or frivolity’. The portrait added up to character assassination, mitigated only by a coy reappraisal in the epilogue. Franklin could not defend herself – she died ten years before The Double Helix appeared – but her former colleague Aaron Klug could and did, at any rate so far as her scientific contributions were concerned. Anne Sayre, a friend of Franklin’s, had meanwhile been preparing a book-length reply to The Double Helix, which appeared in 1975. Sayre’s combative intensity is sometimes heavy-handed, but she deserves credit for radically revising the image created by the ‘Rosy’ of Watson’s account.

The Double Helix was the chief source for a 1987 BBC television film called Life’s Story, in which Franklin was played by Juliet Stevenson. She gave a remarkable performance, and Crick’s impression was that she was ‘not only the true centre of the film – she is almost the only person who really appears to be doing science.’ Here was the beginning of a revaluation of Franklin’s work.

Brenda Maddox goes over much of the same ground as Sayre, but with a lighter touch and a more skilful use of sources. Her emphasis is on Franklin as a scientist, very conscious of her professionalism, her social class, her Jewishness and her abilities. For Maddox’s Franklin there was no small talk, no beating about the bush: she was demanding, belligerent, aloof, at times curt, not an easy person to work with, either at King’s (1951-53) or at Birkbeck (1953-58). But Maddox’s Franklin was also an enthusiast for the outdoors, a mountain climber, immensely kind to those she knew and respected, someone who loved children, gave thoughtful presents and lent her apartment to those in need.

Maddox has made especially good use of Franklin’s scientific and family correspondence, and much about her that seems surprising at first becomes explicable when we read extracts from the letters she wrote to her parents as far back as 1930, when she was ten, and which continue on through World War Two in Cambridge and her first visit to America in 1954. Her privileged upbringing, her education at St Paul’s School for Girls, her family’s response to the anti-semitism of the 1930s, and the pacifists’ readiness to tolerate Hitler, all served to shape her character. There was always a radical streak, sharpened in fierce debates with her conservative father, and revealing an opinionated attitude that found expression in her first impression of America. She was shocked by the ‘overabundance of everything’ and the ‘complete self-confidence of individuals’. She told her brother and sister that most of New England is ‘wasted’. ‘It makes nonsense,’ she added, ‘of all the world-planning talk about cultivating the desert and the jungle – all they have to do is to cultivate America and distribute the products, and world food production problems would be solved.’

When Franklin went to King’s, the plan was for her to study proteins in solution, a project requiring the skills she had developed in earlier work on coal. Subsequently, the director of the unit, John Randall, decided to hand over to her the project that Maurice Wilkins, aided by Raymond Gosling, a doctoral student, had begun – the structural analysis of DNA. Wilkins learned of this only when he returned from overseas to find that his project had been taken from him. The enmity that developed between him and Franklin was thus the result not only of a clash of personalities but of the director’s mismanagement. Randall was right to assign the project to Franklin, who, unlike the rest of the unit’s staff, was experienced in the use of X-ray crystallography (albeit on amorphous substances), but he hadn’t reckoned with the personal repercussions of his decision.

In other respects the outcome was very successful. Within a year, Franklin and Gosling had established the existence of two interconvertible forms of DNA, which they called A and B, the former crystalline, the latter with a higher water content and only semicrystalline. She measured and identified the basic constituent of the crystal of the A form, its ‘unit cell’, and inferred that the water-loving acidic component of the molecule – the phosphate – was on the outside, constituting, along with the sugar component, the backbone of the DNA chain. At this stage, Franklin interpreted these results as meaning that the molecule for both A and B forms was helical. When she sought advice on how to proceed further from her co-workers in Paris she was told to use Patterson analysis on the crystalline A form. Introduced by Lindo Patterson in 1934, this method yields a map of electron density showing the spatial relations of pairs of prominent atoms in the structure. This would allow Franklin to proceed objectively without making unfounded assumptions: the analysis itself would reveal the structure.

Unfortunately, this was not the case. So-called ‘direct methods’, based on accumulated data that give probabilistic predictions of the correct molecular structure, may exist now, but they didn’t in 1951. The fundamental problem is that, unlike rays of light, X-rays cannot be focused to produce an image, or reflected to produce a mirror image. Yet use them we must, if we want to get down to the architecture of these molecules. The patterns of spots and arcs of intensity that were first discovered in 1912 when X-rays were passed through a crystal onto a photographic plate are due to the scattering of the rays from atoms in the crystal. To deduce the arrangement of these atoms from the spots we need to know not only the latter’s intensity and position, but also the phases of the rays that produced them. The ‘phase’ refers to a light-wave’s up or down-ness in its cycle relative to a fixed point – which may, for example, be the atom from which the wave is scattered. The phase problem is that of determining whether the rays producing a given spot had a positive phase (the hill) or negative phase (the valley). Even when the phases are known, if the molecule is large the computations required are enormous. Understandably, the method proved successful only with small molecules or with larger ones whose major structural features had been deduced by other methods – until 1953, that is, when Max Perutz, in Cambridge, used the ‘heavy atom’ technique in haemoglobin to overcome the phase problem. Even then it is misleading to claim, as Maddox does, that the difference in the X-ray diffraction patterns resulting from the replacement ‘reveals the structure’.

To overcome the same problem in DNA by the same method required the synthesis of fragments of DNA, and this wasn’t achieved until a quarter of a century after the publication of the Watson/Crick structure. No wonder that their proposed model was still being challenged as late as 1976. Yet Franklin had been hoping that, when analysed by the Patterson method, her data would reveal the structure. This form of analysis does reveal the principal separations between those atoms in the crystal that scatter the X-rays most strongly, but not which atoms are bonded to which, nor with certainty which separations refer to which parts of the molecular structure or whether they are within the same molecule or between neighbouring molecules. Maddox quotes a crystallographer who thinks of Patterson analysis as ‘the structure speaking to you through those spots’. It can speak, but not unambiguously. Hence the justification for Franklin’s mistaken view in the winter of 1952-53 that one form (A) of the two principal forms of DNA was non-helical.

There were other resources on which she might have drawn, however. First, Fourier theory could be used to predict the kind of diffraction pattern a given molecular structure would yield: Franklin’s remarkable pattern of the B form, obtained in May 1952, followed strikingly that predicted and published by Cochran, Crick and Vand in the same year, and also arrived at independently by A.R. Stokes in Franklin’s lab in 1951. This was to prove a powerful tool for interpreting the diffraction pattern of DNA. Second, the symmetry characteristics of a crystal impose limitations on the kinds of structure that are possible. Franklin’s identification of the A form as face-centred monoclinic strongly implied that the chains in the crystal ran in opposite directions, and the almost hexagonal packing of the molecules in the fibre was suggestive of a cylindrical molecule. These clues did not mean that the molecule had to be helical, but other explanations were unlikely. Third, the structure had to be able to change length rapidly as levels of humidity changed. On this basis, a helical conformation in one form and a non-helical in the other again seemed unlikely. Fourth, accepted distances and angles between the constituent atoms of the structure had to be observed, and there was no better way to ensure this and to explore possible structures than by model-building. Patterson analysis alone did not and could not give the structure.

Early in 1953 Franklin began model-building, and applied helical diffraction theory to the B form. Her Achilles heel was her commitment to the Patterson analysis of the A form, which she hoped to solve without using a priori methods. She cultivated an opposition to helices in order to distance herself from those, including Watson and Wilkins, who wanted to advise her, but she was in any case not yet convinced that the A form was helical. Would it have taken her much longer to come to the conclusion that the study of the relationship between A and B as helices was the way to go? The answer, though speculative, is surely that it would not.

Fortunately, Maddox does not confine her account to the DNA episode, going on instead to show how successful Franklin was in her work on the tobacco mosaic virus. This virus had been crystallised by Wendell Stanley in 1935, and two decades later was shown by Franklin, Klug and Holmes to be a hollow helical cylinder, its single nucleic acid chain buried in a succession of identical protein units. This work, marking the end of Franklin’s short career, has proved of fundamental importance not only in virology but also in the study of the self-assembly of molecular structures in living cells. Had she lived, some of the myths about Franklin would surely have lost their lustre in the face of such achievements.