Franklin, Rosalind Elsie

(b. London, England, 25 July 1920; d. London, 16 April 1958)

physical chemistry, molecular biology.

The daughter of a banking and artistic family previously unconnected with sciences, Rosalind Franklin was a foundation scholar at St. Paul’s Girls’s School, London, from which she won an exhibition to Newnham College, Cambridge, in 1938. After graduating in 1941 she stayed on to investigate gas-phase chromatography under Ronald Norrish. In 1942 she joined the British Coal Utilisation Research Association where, under D. H. Bangham, she applied her expertise in physical chemistry to the problem of the physical structure of coals and carbonized coals. From 1947 to 1950 Franklin worked under Jacques Méring at the Laboratoire Central des Services Chimiques de l’État, Paris, where she developed her skill in X-ray diffraction techniques and applied them to a detailed and illuminating study of carbons and of the structural changes accompanying graphitization. In 1951 she joined Sir John Randall’s Medical Research Council unit at King’s College, London, to apply these techniques to the problems of the structure of DNA, and in 1953 she moved to Birkbeck College, London, to work similarly on the even more exacting problems of virus structure.

At the British Coal Utilisation Research Association, Franklin developed, with Bangham and other workers, a hypothesis of the micellar organization of coals which provided a satisfactory explanation of their absorptive behavior toward liquids and gases and their thermal expansion. From her study of the fine porosity of a range of coals, by measurements of true and apparent densities, Franklin concluded that their structure was best represented by a model with pore constrictions which gave coals the properties of molecular sieves. In Paris she turned her attention to the application of X-ray diffraction methods to the problems of carbon structure and developed a procedure for the detailed interpretation of the diffuse X-ray diagram of carbons. This allowed her to describe the structure in more precise quantitative terms than had been possible, and she made use of it to study in detail the structural changes that accompanied the formation of graphite when these carbons were heated to high temperatures.

In the course of this work Franklin developed a relation between the apparent interlayer spacing of the partially graphitized carbons and the proportion of disoriented layers, which has proved of considerable value in the industrial study of carbons. In addition, by studying the changes in structure that chars of different origin underwent on heating, she established that there are two distinct classes of carbons—those which form graphite on heating to high temperatures (the graphitizing carbons) and those which do not (the nongraphitizing carbons)— and related these differences in behavior to structural differences in the parent chars. She showed, in particular, that the graphitizability increases with the finestructure porosity and this, in turn, she believed to be related to the cross-linking between the crystallites.

Franklin’s work on coals brought her into contact with Charles Coulson, through whom she was introduced to Randall, and with the award of a Turner Newall Fellowship she went to work in the King’s College Medical Research Council Biophysics Unit. At that time (January 1951) Raymond Gosling, under M. H. F. Wilkins’ direction, had obtained diffraction pictures of DNA showing a high degree of crystallinity; sharper pictures were obtained with higher ambient humidity.

Franklin and Gosling conducted a systematic study of the effect of humidity on the X-ray pattern produced. Using salt solutions to control humidity, they showed that there are two distinct intramolecular patterns, which they found to be producible from the same specimen: the crystalline “A” pattern at 75 percent relative humidity and a new “Wet” paracrystalline pattern at 95 percent relative humidity. In a report which Franklin gave on this work in November 1951, she described this discovery and went on to show, as Wilkins had a year before in Cambridge, that the patterns were consistent with a helical conformation. She discussed how the A ?B transformation takes place and suggested, quite correctly, that the phosphates are on the outside of the helices and in the “A” form are held parallel to each other by electrostatic attraction between O^- and Na⁺. When water is added, it penetrates between the helices, thus destroying the electrostatic attraction which holds them in parallel alignment. She said little about the forces operating inside the helices but mentioned hydrogen bonding between keto and amino groups of the bases.

Despite this promising beginning Franklin was too professional a crystallographer to proceed further in this way. Instead, she thought to solve the structure of DNA in an inductive manner by using Patterson functions and superposition. While publicly she heaped scorn on those who were convinced that DNA is helical, in her unpublished reports she stated that such a conformation is probable for the B form and not inconsistent with the A form. A spurious case of double orientation encountered in April 1952, which when indexed showed marked radial asymmetry (all left-hand reflections were indexed hKL and all righthand ones hkl?), led her to seek nonhelical structures for the A form. Earlier ambiguities in the indexing of the A diagram had led Franklin to embark on a Patterson analysis. This helped her to obtain accurate parameters for the unit cell. Yet the cylindrical Patterson function obtained by Gosling in July 1952 strengthened her antihelical views, although the arrangement of peaks was consistent with a helix. She was misled, by what appeared to be clear evidence of a structural repeat at half the height of the unit cell, into ruling out helices for the A form, since no DNA chain could possibly be folded into a helix with a pitch equal to half the height of the unit cell (fourteen Å.)

At this time Franklin was thinking in terms of antiparallel rods in pairs back-to-back, forming a double sheet structure. Then she investigated diagonal rod structures such as would simulate the diffraction pattern of a helix, but by January 1953, when she started model building, she found such structures impossible to build. Still rejecting single- or multistrand helices, she investigated a figure-eight structure in which a single chain formed a long column of repeating eights. This, she believed, would account for the halving of the unit cell in the cylindrical Patterson function and clearly provided a form of tight packing which could be unfolded to give the dramatic increase in length (30 percent) when structure A changes to structure B. She knew that the helix in the extended B form is close-packed and was doubtful that the same type of structure could pack down even more densely.

At the end of February 1953 Franklin turned to the B Pattern, and for two weeks she weighed the merits of single and multiple helices. In a paper dated 17 March which she wrote with Gosling, she ruled out triple-strand and equally spaced double-strand helices and stated that “if there are two nonequivalent, i.e., unequally spaced coaxial chains these are separated by 3/8th. of the fibre axis period.” This is the conformation of the sugar-phosphate backbones as found in the Watson-Crick model. On the following day Franklin returned to the Patterson function of the A form, only to learn that Watson and Crick had solved the structure of the B form. She and Gosling quickly expanded and rearranged their draft paper of 17 March in the light of the Cambridge discovery so that it could appear in the 25 April issue of Nature, which contains Watson and Crick’s paper on their model.

Franklin deserves credit for having discovered the A ?B transformation and characterized the diffraction patterns of these forms of DNA; for providing Watson and Crick with vital data, in particular the parameters of the unit cell; for exposing the errors in their first unpublished model; and for marshaling the evidence in favor of the phosphates being on the outside of the helix. It was also she who, with the aid of the special tilting camera built by Gosling, discovered the meridional reflection on the eleventh layer line in the A pattern and was the first to show how the B form can pack down more tightly to give the A form with eleven residues in one turn of the helix. Although she had been misled by the cylindrical Patterson function, this did provide the most refined evidence in favor of the Watson-Crick model at the time of its discovery in 1953. Franklin and Gosling’s rarely cited paper on this subject appeared in Nature on 25 July 1953.

For the last five years of her life Franklin worked in the Crystallography Laboratory of Birkbeck College, London, supported first by the Agricultural Research Council and later by the U.S. Department of Health. There she continued to publish on her earlier work on coals, completed the writing up of her DNA work, and took up the structure of tobacco mosaic virus (TMV). By 1956 she had greatly improved on J. D. Watson’s X-ray pictures of 1954. With the aid of material supplied by Heinz Fraenkel-Conrat and by Gerhard Schramm, Franklin and her co-workers, A. Klug and K. C. Holmes, were able to reject the picture of TMV as a solid cylinder with the RNA in the center and the protein subunits, possibly of two types, on the outside. They showed that the particles are hollow, that the protein subunits are structurally of one type only, and that forty-nine such units are packed in helical array around the axis in the axial repeat period of sixty-nine Å. The greatest achievement of the Franklin team was the location of the RNA helix embedded within the protein fraction at a radial distance of forty Å. from the axis. From a study of the X-ray diagram of TMV, Franklin and Klug resolved the discrepancy between estimates of the maximum radius and the packing radius by postulating the morphology of the protein as “a helical array of knobs, one knob for each sub-unit.” Shortly before her death from cancer, Franklin instituted work which was later to justify her conclusion that the RNA in TMV is present in the form of a single-strand helix.

Franklin was a deft experimentalist, keenly observant and with immense capacity for taking pains. As a result she was able with difficult material to achieve a remarkable standard of resolution in her X-ray diagrams. Although a bold experimentalist, she was critical of speculation, favoring an inductive approach which proved very successful in her work on coals and TMV but which allowed others to get ahead of her in her work on DNA. Where those with a more intuitive approach rejected antihelical data as spurious, Franklin felt obliged to invent other conformations which might yield a helical-type pattern. She would not trust to the principle of exclusion, nor was she confident of the “Obvious” deductions dictated by physical intuition. Hence her work was not marked by great originality of thought. Her theory of graphitization, for instance, although the best of its day, was traditional in character and belongs to what is now regarded as the “Classical” period. Her great strength lay in her technical innovations and her employment of precise technique on difficult macromolecules. When she died at the age of thirty-seven, she had won international recognition both as an industrial chemist and as a molecular biologist.

BIBLIOGRAPHY

I. Original Works. Franklin and her co-workers published about forty papers. A complete list of her publications on the structure of viruses is in her paper written with D. L. D. Casper and A. Klug, “The Structure of Viruses as Determined by X-Ray Diffraction,” in C. S. Holton et al., eds., Plant Pathology: problems and progress, 1908–1958 (Madison, Wis., 1959), pp.447–461. This paper provides a broad review of Franklin’s work and contains a tribute to her by W. M. Stanley.

Of her other papers the following appeared in Acta crystallographica: “The Interpretation of Diffuse X-Ray Diagrams of Carbon,” 3 (1950), 107–117; “A Rapid Approximate Method for Correcting Low-Angle Scattering Measurements for the Influence of the Finite Height of the X-Ray Beam,” ibid., 158–159; “The Structure of Graphitic Carbons,” 4 (1951), 253–261; “The a Dimension of Graphite,” ibid., 561; and “The structure of Sodium Thymonucleate Fibres. I, II, & III,” 6 (1953), 673–677, 678–685; 8 (1955), 151–156.

Other important papers are “A Note on the True Density, Chemical Composition and Structure of Coals and Carbonized Coals,” in Fuel, 27 (1948), 46–49; “A Study of the Fine-Structure of Carbonaceous Solids by Measurements of True and Apparent Densities. Part I. Coals. Part II. Carbonized Coals,” in Transactions of the Faraday Society, 45 (1949), 274–286, 668–682; “Crystallite Growth in Graphitizing and Nongraphitizing Carbons,” in Proceedings of the Royal Society, 209A (1951), 196–218; “Molecular Configuration in Sodium Thymonucleate,” in Nature, 171 (1953), 740–741; and “Evidence for 2-Chain Helix in Crystalline Structure of Sodium Deoxyribonucleate,” ibid., 172 (1953), 156–157.

Her last paper, written with A. Klug, was published after her death: “Order-Disorder Transitions in Structures Containing Helical Molecules,” in Discussions of the Faraday Society, 25 (1958), 104–110.

II. Secondary Ltterature. Obituary notices appeared in The Times (19 Apr. 1958), p.3; and Nature, 182 (1958), 154.

For two very different accounts of Franklin’s work on DNA, see J. D. Watson, The Double Helix (New York, 1968), Passim; and A. Klug, “Rosalind Franklin and the Discovery of the Structure of DNA,” in Nature, 219 (1968), 808–810, 843–844; corrigenda, 879, 1192; correspondence, 880.

Robert Olby