Klug dna

HISTORICAL PERSPECTIVE

The Discovery of the DNA Double Helix

Aaron Klug

MRC Laboratory of Molecular Biology, Hills Road, Cambridge CB2 2QH, UK

Introduction

Fifty years ago, on 25th April 1953, thereappeared three papers in the journal, Nature,which changed our view of the world. The struc-ture of the DNA double helix, with its complemen-tary base-pairing, is one of the greatest discoveriesin biology in the 20th Century. It was also mostdramatic, since, quite unexpectedly, the structureitself pointed to the way in which a DNA moleculemight replicate itself, and hence revealed the“secret of life”. The structure was solved in theCavendish Laboratory, Cambridge by FrancisCrick and James Watson, using X-ray diffractiondata from fibres of DNA obtained by RosalindFranklin at King’s College, London.

This article aims to tell the story of how thiscame to happen: the origin of the research onDNA, the early investigations by Maurice Wilkinsat King’s College, the sorting out of the two formsof DNA by Franklin, the wrong paths taken, theintervention of old rivalries from an earlier gener-ation (Lawrence Bragg and Linus Pauling), andthe final model-building by Watson and Crick togive the three dimensional structure.

I will also describe the initial, mostly hesitant,reception of the proposed structure, and its confir-mation by biochemistry by Arthur Kornberg andby X-ray crystallography at King’s College byWilkins’ group. Yet this remained a discovery inchemistry, until the biological principle of “semi-conservative” replication was proved by Messelsonand Stahl in 1958.

The transforming principle

In 1945 The Royal Society of London awarded itshighest honour, the Copley Medal, to OswaldAvery of the Rockefeller Institute of New York for“establishing the chemical nature of the transform-ing principle”.

The transforming principle was an extract bymeans of which a non-pathogenic mutant of thepneumococcus bacterium could be transformedinto a pathogenic form. The President of theSociety, Sir Henry Dale, commented “Here surelyis a change to which, if we were dealing withhigher organisms, we should accord the status ofa genetic variation, and the substance inducingit—the gene, one is tempted to call it—appears tobe a nucleic acid of the desoxyribose type. What-ever it be, it is something which should be capableof complete description in terms of structuralchemistry”. The hesitation about “gene” reflectsthe belief then held by some biochemists and biol-ogists that bacteria did not possess genes.

Eight years later the President’s challenge wasanswered: there was a complete description of the3D structure of DNA—what a chemist would callits configuration—the double helix by Watson andCrick (Figure 1), using X-ray diffraction data fromFranklin and Wilkins. This paper aims to describehow this came about. Much of the story has beentold in parts, but Franklin’s scientific work hasnever been fully described, and I have thereforedrawn on her notebooks, now in the Archives ofChurchill College, Cambridge, to document it.

We begin with chemical structure of DNA, thatis, how the links in phosphate–deoxyribose sugarbackbone are made and how the heterocyclicnitrogenous bases are connected to the sugars.This had been worked out only two years earlierby Brown and Todd in Cambridge (Figure 2).

The structure of DNA

Since the structure of DNA is so well knownthere is little point in keeping it to the end as adenouement of the story.

The double helix (Figure 3) consists of two inter-twined helical phosphate-sugar backbones, withthe heterocyclic DNA bases projecting inwardsfrom each of the two strands. The two chains areantiparallel, running in opposite directions, andare related by a 2-fold axis of symmetry (dyad)

0022-2836/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.

E-mail address of the corresponding author:[email protected]

doi:10.1016/j.jmb.2003.11.015 J. Mol. Biol. (2004) 335, 3–26

perpendicular to the axis of the double helix. Thebases are arranged in purine-pyrimidine pairs,adenine with thymine, guanine with cytosine,linked by hydrogen bonds (Figure 4), and thesebase-pairs are stacked on top of each other alongthe helix axis at a distance of 3.4 A apart. The gly-cosidic bonds (the links between sugar and base)are related by the perpendicular dyad, so thatthey occur in identical orientations with respect tothe helix axis. The two glycosidic bonds of a pair

will not only be the same distance apart for allpairs, but can be fitted into the structure eitherway round. This feature allows all four bases tooccur on both chains, and so any sequence ofbases can fit into the double helix.

The two chains are said to bear a complementaryrelationship to each other. This means, as Crick andWatson spelt out in their second paper in Nature inMay 1953, that when the two chains come apartduring replication of DNA, each can be used as atemplate to assemble a duplicate of its former part-ner (Figure 5). The crucial feature of the structureof DNA is not therefore the actual double helicalform of the two phosphate-sugar chains—eye-catching as it is—but the unique pairing of thebases projecting from each strand.

Structural research on DNA

In 1945, most biochemists, had doubts whethersomething as simple as what DNA was thought tobe—repeats of the four nucleotide bases—could bethe genetic substance. More complex moleculeslike proteins—chromosomal proteins—werethought to be more likely candidates.

There were some who did believe in DNA, inparticular, the “phage group” in the USA led byMax Delbruck and Salvador Luria. This group,mostly geneticists, studied bacterial viruses, bac-teriophages. A younger member of that group wasJames Watson, who in October 1950 went toCopenhagen to learn nucleic acid chemistry, butwas converted to a structural approach by hearingMaurice Wilkins speak at a Conference on LargeMolecules in Naples in May 1951. Wilkinsdescribed his X-ray diffraction studies on fibres ofDNA and showed a diffraction pattern with muchmore detail than had been obtained by earlierworkers, Astbury and Bell in 1938. Moreover, theyindicated a degree of crystallinity which raised thepossibility of a molecular interpretation by X-rayanalysis.

Watson therefore decided to go to a laboratorywhere he might learn X-ray diffraction techniques,and, failing to interest Wilkins, he eventuallymoved his fellowship to the MRC Unit in Cam-bridge headed by Max Perutz. Here the structuresof the proteins haemoglobin and myoglobin werebeing tackled. Watson arrived in September 1951and met Francis Crick, who was working for hisPhD on haemoglobin under Max Perutz, andfound him like-minded about the importance ofDNA.

A preamble on X-ray diffraction bycrystals and by fibres

X-ray crystallography provides a way of dedu-cing the structure of a molecule by analysing thediffraction pattern produced when a beam ofX-rays falls on a crystal in which the molecules

Figure 1. Space-filling atomic model of the DNAdouble helix. Colouring: phosphorus yellow; oxygenred; carbon dark blue; nitrogen light blue; hydrogenwhite.

4 The Discovery of the DNA Double Helix

are regularly arranged in three dimensions (Figure6). The pattern is nothing like a conventionalphotograph: it shows a set of spots of varyingintensity and inferring the structure from the pat-tern is not a direct process. This is because eachspot corresponds to a diffracted wave from themolecules lying in a particular set of planes in thecrystal. The molecular structure of the crystalcould be reconstructed mathematically from aknowledge of the amplitudes and phases of thediffracted waves—amplitude means strength ofthe wave (which is measurable from the spot inten-sity); and phase means the positions of the peaksand troughs of the wave relative to some referencepoint, but the phase is lost in the recording. Hencearises the so called phase problem in X-ray crystal-lography which is to develop methods for deter-mining indirectly these lost phases. For smallmolecules, analytical methods have been devel-oped, and for large molecules like proteins the pro-blem was solved in 1953 by Max Perutz by hisimplementation of the heavy atom isomorphousreplacement method.

Fibrous macromolecules—polymers of smallunits, the monomers, regularly (or “equivalently”)arranged—present a further challenge in X-ray dif-fraction, since in fibres, the long molecules, thoughroughly parallel to one another, are usually not allrotationally oriented relative to one another in aregular manner. The observed diffraction patternthen represents the rotational average of the pat-terns that would be given by different orientations.If the chemical structure of the monomer is known,as was, for example, the case of rubber or cellulose,

Figure 2. The chemical formula ofa chain of a DNA molecule (DMBrown and AR Todd, 1952 J. Chem.Soc. p. 52). The backbone is madeup of alternating sugar (2-deoxyri-bose) and phosphate groups. Eachsugar has attached to it a side-group by a glycosidic linkage. Theside groups consist of either a pur-ine base (adenine or guanine) or apyrimidine base (cytosine or thy-mine). Note that the backbone hasa directionality because the phos-phate group is linked differently tothe sugars on either side (to the 30

carbon atom of one sugar and tothe 50 carbon atom of the other). Aphosphate-sugar linked to a base iscalled a nucleotide. The DNA chainis synthesised from such nucleo-tides in the 50 –30 direction.

Figure 3. Schematic illustration of the DNA doublehelix as later sketched by James Watson (“The DoubleHelix”, 1968).

The Discovery of the DNA Double Helix 5

Figure 4. The pairing of bases byhydrogen bonds: adenine with thy-mine, guanine with cytosine (Crickand Watson, 1954 Proc., Roy. Soc. A223 80–96, who showed only twohydrogen bonds for the G:C pair,though tentatively suggesting athird, later confirmed by Pauling).

Figure 5. Principle of replicationof the double helix. The helixunzips and each chain acts as atemplate for the synthesis of acomplementary chain, thus creatingtwo double helices, which are iden-tical copies of the first. DATP,dTTP, adenine and thymine nucleo-tide triphosphates, (energy-rich)monomers being incorporated atthe next step in the growing chains.


then the polymer structure can be solved, by build-ing models and comparing the calculated diffractionpatterns with the observed ones. This is the modelbuilding approach, which was used by Watsonand Crick for DNA. The problem is that there isrotation about the single chemical bonds betweenmonomers (and also usually within them), soother constraints must be used to fix how themonomers join head to tail.

King’s College, London

Wilkins was a senior member of the MRC Bio-physics Unit at King’s College, London, set up by

(Sir) John Randall in 1946 after the War to carryout “an interdisciplinary attack on the secrets ofchromosomes and their environment”. Wilkinsworked to develop special microscopes, but havingheard of the greatly improved methods devised byRudolf Signer at Berne for extracting long unbro-ken molecules of DNA, he obtained some of thematerial and found a way of drawing uniformfibres from a viscous solution of DNA (Figure 7).Examination under polarized light showed themto be well ordered, characteristic of long moleculesoriented parallel to one another. He enlisted thehelp of a graduate student in the Unit, RaymondGosling, who was studying ram sperm by X-raydiffraction. By keeping the fibres in a wet

Figure 6. Schematic diagram ofX-ray diffraction by a crystal con-taining a regular 3D arrangementof molecules. The pattern of dif-fracted waves depends on the par-ticular setting of the crystal relativeto the incident X-ray beam. A full3D set of X-ray data is collected byrotating the crystal into differentsettings.

Figure 7. A fibre drawn from aDNA gel after exposure to an X-raybeam which has punched a hole init (MHF Wilkins, WE Seeds and RGGosling Nature 1951, 167, 759),reproduced from Gosling’s PhDthesis, King’s College, London 1954.


atmosphere, Gosling and Wilkins obtained the X-ray diffraction photograph that Wilkins latershowed at Naples and which so excited Jim Wat-son (Figure 8). Other early diffraction photographsof various specimens (Figure 9) showed hazy pat-terns, later understood to be indicating helical fea-tures (Figure 10).

Rosalind Franklin

In January 1951, the King’s College group wasstrengthened by the arrival of Rosalind Franklin(Figure 11). She was a physical chemist, who wasthen studying the structure of carbons (cokes andchars) using X-ray diffraction methods. She hadbeen at a CNRS lab in Paris, where she learned,and improved, X-ray diffraction techniques fordealing quantitatively with substances of limitedinternal order. These presented much more diffi-culty than the highly ordered crystals which X-raycrystallographers were using to solve the struc-tures of small molecules. It is important to realisein what follows that, in Paris, Franklin gained noexperience of such formal X-ray crystallography.

The combination of these X-ray diffraction tech-niques and chemical preparatory skill attractedthe attention of Randall, and Franklin was invitedby him to bring her experience to London. Ran-dall’s purpose was clearly to put more professionaleffort into the DNA work begun by Wilkins andGosling. Randall, however, left an unfortunateambiguity about the respective positions of Wilkinsand Franklin, which later led to dissensionbetween them about the demarcation of the DNAresearch at King’s. To this must be added the verydifferent personalities of the two. A letter of Ran-dall to Franklin in December 1950 (Figure 12)makes it clear that “on the experimental X-rayeffort there would be for the moment only yourselfand Gosling”. Wilkins did not see this letter, andwas away when Franklin arrived in January 1951and Gosling was formally placed under her super-vision. Nevertheless he still apparently thought ofFranklin as a member of his team.

Wilkins handed over the Signer DNA to them,and turned to an X-ray study of sperm whereDNA is complexed with proteins. It should beremembered that at the time, no one, not even Wat-son, had imagined that the 3D structure of DNAalone, important as that might turn out to be,would by itself indicate how the molecule repli-cated itself, and hence reveal “the secret of life”.

Within the first year Franklin transformed thestate of the field. By drawing thinner fibres she

Figure 8. The first clear crystalline pattern from aDNA fibre, King’s College 1950, in what was later calledthe A form (from RG Gosling “Genesis of a Discovery:DNA Structure”, ed S Chomet 1993). This was shownby Maurice Wilkins at the Naples conference attendedby James Watson (courtesy, R G Gosling).

Figure 9. Early X-ray diffraction patterns obtained by the King’s College group, about 1950, suggestion of a helicalstructure (cf. AR Stokes, in “Genesis of a Discovery: DNA Structure”, ed S. Chomet, Newman Hemisphere Press 1993).


was able to enhance the alignment of the DNAmolecules within the specimen, and these speci-mens, together with finer collimation of the X-raybeam generated from a microfocus X-ray tubewhich she and Gosling had assembled, producedsharper diffraction patterns (Figure 13). These,however, showed variable features, and it was notuntil Franklin made a systematic study of thefibres, that the problem was solved.

The A and B forms of DNA

In a crucial advance, Franklin controlled the rela-tive humidity in the camera chamber by using aseries of saturated salt solutions and thus was ableto regulate the water content of the fibre speci-mens. In this way she showed that, depending onthe humidity, two forms of the DNA moleculeexisted, which she later named A and B, and

Figure 10. Diffraction patternproduced by a continuous helix: anoptical analogue (from KC Holmesand DM Blow. “The Use of X-rayDiffraction in the Study of Proteinand Nucleic Acid Structure”, Inter-science, 1966). Note the X-shapedfan of reflections emanating fromthe origin (at the centre).

Figure 11. Rosalind Franklin onholiday in Italy in the summer of1950 (photo from V Luzzati, Paris).

Figure 12. Excerpt from a letterdated 4 December 1950 from Pro-fessor JT Randall, King’s College,London, to Rosalind Franklin inParis (Franklin papers, ChurchillCollege Archives, Cambridge).


defined the conditions for the transition betweenthem (Figure 14). The A form, which she firstcalled “crystalline”, is found at, and just below,75% relative humidity. Above that point there isan abrupt transition to the B form, which she orig-inally called “wet”. The X-ray patterns of the Aand B forms are shown in Figures 15 and 16,respectively.

It became clear that all previous workers hadbeen working, unbeknown to themselves, mostlywith a mixture of the two forms, or at best withpoorly oriented specimens of the A form, and, inretrospect, with occasionally hazy pictures of theB form.

The B form pattern illustrated in Figure 16, is thesuperb picture B51, which Franklin obtained laterin May 1952, and which has achieved iconic status.It was this picture that was shown by Wilkins toWatson in early 1953, and prompted the Cam-bridge pair into active model building. But evenless striking X-ray patterns, which Franklin hadobtained by September 1951 (Franklin and Gosling1953, Acta Crystallographica, Figure 2), showedclear evidence of a helical structure. The theory ofdiffraction by a helix had been worked out byAlex Stokes at King’s at the behest of Wilkins(unpublished, 1951), and also independently byCochran, Crick and Vand the same year, publishedin early 1952. The characteristic feature of the pat-tern is the X-shaped pattern of streaks arranged ina set of layer lines, from which it can be deducedthat the pitch of the helix in the B-form is 34 A.

A strong X-ray reflection lies on the meridian,corresponding to a spacing of 3.4 A: this is pro-duced by the regular stacking of the bases on topof each other. Since the helix pitch is 34 A, thismeans that the helix, whatever it is in detail,repeats after 10 ( ¼ 34/3.4) units per turn. Thisphotograph is particularly striking in that itshows, not only the X shaped pattern of streaks inthe centre, but also secondary fans emanatingfrom the two 3.4 A meridional reflections, top andbottom, and running obliquely to the equator.These are characteristic of a discontinuous helix(Figure 17), as is to be expected from the discretemoieties in a phosphate-sugar chain.

In the A form, the repeat of the structure is 28 Acompared with 34 A in the B, consistent with amacroscopic shrinkage of 25% in the lengths ofthe fibres. The A form does not show the character-istic X, but there is a gap on the meridian of thephotograph, consistent with a helical structure asFranklin recognised (and described in November1951—see below). The A form, as later recognisedby Watson, is a somewhat more tightly woundform of the B double helix, in which the baseschange their tilt obliquely to the fibre axis (Figure18), thus obscuring the characteristic X-shaped fanof reflections expected from a simple helicalstructure.

For, despite her discovery of the simpler B pat-tern, Franklin at first directed her attention mostlyto the A form. Here, the molecules themselves arenot in random rotational orientations, as in B, but

Figure 13. X-ray diffraction pat-tern of a DNA fibre 1951, laterunderstood to be a “mixture” ofthe A and B forms of DNA (Gosl-ing, PhD thesis, London 1954).


packed regularly in small crystallites in a crystallattice (Figure 19). However, the crystallites arerandomly oriented so that the 3D X-ray data isscrambled into two dimensions on the photo-graphic plate. The X-ray pattern nevertheless stillshows sharp “spots” and offered the possibility of

an objective crystallographic analysis because ofthe greater wealth and precision of the 3D diffrac-tion data which could be extracted from the 2Dpattern.

Figure 14. Franklin’s measure-ments of water uptake by DNAfibres, made by weighing them atdifferent ambient relative humid-ities, the latter controlled by theuse of saturated salt solutions.(Franklin papers, Churchill CollegeArchives, Cambridge; hithertounpublished, described qualitat-ively in Franklin and Gosling, ActaCrystallog. 1953 6, 673). There issome hysteresis in that the curveson re-drying do not follow exactlythe initial wetting curves. There isa marked transition at about 75%relative humidity, when the DNAstructure changes abruptly fromthe “crystalline” form (later calledA) to the “wet” form B. Annotationin Franklin’s notebook by theauthor A.K.

Figure 15. X-ray diffraction pattern of the A form ofDNA (Franklin and Gosling, Acta Crystallog. 1953, 6,673, Figure 1.

Figure 16. X-ray diffraction pattern of the B form ofDNA (Franklin and Gosling, Acta Crystallog. 1953, 6,673, Figure 4; The B form pattern is that reproduced inFranklin and Gosling Nature 25 April 1953, 171, 740,also known as photograph B51.


In retrospect this was a misjudgement, but it wasa reasonable decision at the time, because, if cor-rectly interpreted, the A pattern would yield moreprecise information about the DNA molecule. Shedecided to use what is called Patterson functionanalysis on the X-ray data she had measured onthe A patterns, and, as Gosling said later, let the

data speak for itself. This Patterson method is anindirect method, which had been used at higherresolution to solve the structures of small mol-ecules, but never for such large unit cells.

Franklin’s Colloquium, November 1951:Watson and Crick’s first model

In November 1951 Franklin gave a colloquiumon her work at King’s College which Watsonattended. There was much contact on and offbetween Wilkins and Crick, who were friends,and this led to several visits by Watson to King’s.

The draft of Franklin’s colloquium and heraccompanying notes survive in the Archives ofChurchill College, Cambridge. She describes a[very] dry form (1) and, the two forms “crystalline”(2) (later A) and “wet” (3) (later B) which is noteasily re-wetted. She gives the crystal parameters,and the lattice symmetry (monoclinic space groupC2), and also the density of A, from which shededuced that there were two or three chains ofDNA per lattice point. The packing is pseudo-hex-agonal, which implies that the molecules have anapproximately cylindrical shape with a diameterof about 20 A, Her notes read: “Evidence for spiralstructure [we would now say helical]. Straightchain untwisted is highly improbable. Absence ofreflections on meridian in xtalline form suggests

Figure 17. Diffraction pattern given by a discontinuous helix made of discrete units (right), compared with thatgiven by a continuous helix (left, cf. Figure 10 above). In this example, there are five units per turn of the helix, givingrise to a meridional reflexion (i.e. on the axis) on the fifth layer line. There is a subsidiary X-shaped fan, emanatingfrom the meridional reflexion, as well as the main fan emanating from the centre of the pattern, This can be seen inFranklin’s X-ray pattern of the B form (Figure. 16).

Figure 18. Early diagrams of the structures of the Aand B forms of DNA (GB Sutherland and M Tsuboi,Proc. Roy. Soc. A 1957, 239, 446, the A form after Wilkinset al. Nature, October 1953, 172, p. 759).


spiral structure… Nucleotides in equivalent pos-itions occur only at intervals of 27 A [correspond-ing to] the length of turn of the spiral”.

On the basis of the above, Franklin put forwardher view that the molecular structure in the Aform was likely to be a helical bundle of two orthree chains, with the phosphate groups on theoutside. The bundles are separated by weak linksproduced by sodium ions and water molecules(Figure 20). At the higher humidity of the B form,a water sheath disrupts the relationship betweenneighbouring helical bundles, and only the paralle-lism of their axes is preserved. (The same con-clusions are found in Franklin’s Fellowship Reportfor the year ending 1951). Watson (and others)have stated in their reminiscences that Franklindid not mention the B form, but her draft is quiteexplicit about the helical bundle being preservedin the transition from A to B. (Indeed, her notesread “Helical structure in the [wet] form cannot be

the same as in the [crystalline] because of largeincrease in length”.)

Watson took the news—as little, or as much, ashe understood of it—back to Crick in Cambridge,and, now with some structural information tohand, they decided to build a model. They hadurged this approach on the King’s group, butreceiving no response, now felt justified in attempt-ing this themselves. The King’s group was invitedto see the result—a model built in a week. Themodel was of three helical chains with the phos-phates on the inside, neutralised by cations, withthe bases pointing outwards. Franklin askedwhere was the water, and received the reply thatthere was not any. It turned out that Watson, notunderstanding the relationship between a unit cellof a crystal and the asymmetric unit, had conveyedthe wrong water content. After this debacle, SirLawrence Bragg, the head of the Cavendish Lab-oratory, firmly vetoed any further work on DNA

Figure 19. The packing of DNAmolecules in the A and B form com-pared with those in a perfect crys-tal. The diagrams show schematiccross-sections of the arrangements.

Figure 20. Diagram from Frank-lin’s notes for the Colloquium shegave at King’s College in November1951, annotated by A.K. The DNAmolecules in the A form are rep-resented as helical bundles of two,or three, chains (here two), withthe bases in the inside, the phos-phates on the outside, and the indi-vidual molecules associated in thefibre through water and ionic links(dotted lines). Each molecule hassix near neighbours, four equiva-lently related and two othersapproximately related. (Franklinpapers, Churchill CollegeArchives).


at the MRC Unit in Cambridge. In future it wouldbe done solely at the Unit at King’s College.

Non-helical DNA?

Franklin pressed ahead with the Patterson anal-ysis of the A Form. There is no question that allalong she held the view that B form was helical(Figure 21), but could not see a way to solve itexcept by model building, a path she was reluctantto follow. She knew of Pauling’s success in 1951 inpredicting, by model building, the a-helical and b-sheet configurations of the polypeptide chains ofproteins, but she equally well knew of the contem-porary failure of Bragg, Kendrew and Perutz onthe same problem—the “greatest fiasco of myscientific life”, Bragg later called it. This last deba-cle of Watson and Crick would only have con-firmed her decision to avoid model-building andrather to try an analytic crystallographic approachon the A form.

However, an unfortunate mechanical accident inone of the specimens led Franklin to take a wrongpath. In the spring of 1952, one DNA fibre gave anX-ray pattern showing strong “double orientation”,

that is, the 3D crystallites in the A form were not allin random orientation about the fibre axis, butsome orientations occurred move frequently thanothers. This suggested to her that the symmetry ofthe crystallite was far from cylindrical, whichmight rule out a helical structure in the A form.Franklin concluded that this possibility had to beconsidered. It is this view of hers which gave riseto her supposed “anti-helical” stance, but for herit was a question which had to be answered.Unwisely, she ignored Crick’s remark to her, madein a tea-queue at a meeting, that the double orien-tation was an accident to be dismissed.

In fact she seems to have persuaded Wilkins,even though relations were strained betweenthem, to the same view. Thus, ironically, whileFranklin does not mention this in a Report writtenin late 1952 for an MRC Subcommittee on thework of the King’s Unit, Wilkins does so, andaccepts the possibility of a non-helical interpret-ation of the A form. This Report is the MRC Sub-Committee Report, which gave crucial informationto Watson and Crick in February 1953 for the build-ing of their correct model of DNA (see below).

This misjudgement on Franklin’s part influencedher attempts to interpret the Patterson map of the

Figure 21. Excerpt from Frank-lin’s notebook May 1952, showingan analysis of the X-ray pattern 49B(a precursor of the famous B51picture, Figure. 16 above) in termsof helical diffraction theory (Frank-lin papers). This was done at atime when she was questioningwhether the A form was helical.


A form. She sought explanations in terms of rodsor sheets, or a “figure of eight”, all of which natu-rally failed. She was apparently thinking of the Aform as an unwound version of the helices in theB state (rather, I imagine, like the b-sheet structureis to the a-helix in proteins and polypeptides). Pre-sumably she thought the A-to-B transition a pro-found change of structure, because she notes,more than once, that, during the transition, the spe-cimen fell off the end of the X-ray collimator towhich it was attached.

One correct result which emerged in January1953, from her application of the so-called super-position method to the Patterson map, was thatthe A form contained two chains, and that theyran in opposite directions (Figure 22). Had shebeen a crystallographer, and understood the mean-ing of the crystal symmetry, C2 face centred mono-clinic, which she herself had established muchearlier, she could have deduced this result at once.Of all the protagonists in the story, only Crick

understood this. Moreover, C2 was the spacegroup symmetry of the ox-haemoglobin crystalswhich he was studying for his PhD. It meant that,if the A structure was helical, it would consist oftwo chains, or strands, running in opposite direc-tions, related by a 2-fold axis of symmetry perpen-dicular to the fibre axis, and hence to the pair ofchains. (Franklin hardly ever reminisced aboutDNA in the years I worked with her on virus struc-ture at Birkbeck College, but she once said that shecould have kicked herself for missing the impli-cations of the C2 symmetry).

Franklin’s Patterson analysis ran into animpasse, and in early February, she turned to herB-form, the X-ray pattern which was clearlycharacteristic of some kind of helical structure(Figure 23). Her notebooks show her shuttlingback and forth between the two forms. She had bynow abandoned her attempts to interpret the Aform in non-helical terms. On the 23rd February(Figure 24) she writes “If single-strand helix as

Figure 23. Excerpt from Frank-lin’s notebook, 10 February 1953(Franklin papers). Annotations bythe author A.K. Franklin returns tothe question of the number ofchains in the B form.

Figure 22. Excerpt from Frank-lin’s notebook on 19 January 1953(Franklin papers). Annotations bythe author A.K. Franklin deducesby Patterson superposition analysisthat the A form contains two chainsrelated by a 2-fold axis of symmetry(the oval symbol). Also are notedChargaff’s base ratios, and Broom-head’s crystal structures of adenineand guanine, which showed thecorrect tautomeric forms.


above is the basis of structure B, then Structure A isprobably similar, with P–P distance along fibreaxis ,3.4 A, probably 2–2.5 A”. On the 24th Febru-ary (Figure 25) she is at last making the correct con-nection between the A and B forms—both havetwo chains.

Of course, she had no idea that, at that very timein Cambridge, in February 1953, Crick and Watsonwere now back to model building of DNA. Norwere they aware of what Franklin had beendoing—Watson wrote later in his book “TheDouble Helix” that Franklin’s instant acceptanceon first seeing their model surprised him. He hadthen no idea how close she had come to it.

By March 1953, using helical diffraction theory,Franklin had carried the quantitative analysis ofher B form patterns to the point where the pathsof the backbone chains were determined. She hadmoved to Birkbeck College to J D Bernal’s Depart-ment of Physics on14th March and there shewrote up her work in a typescript dated 17thMarch, that is, one day before the manuscript ofWatson and Crick’s structure, prepared for Nature,reached King’s. Franklin’s draft (Figure 26) con-tains all the essentials of her later paper (with Gosl-ing) in Nature in April, which, together with oneby Wilkins, Stokes and Wilson, accompanied Crickand Watson’s paper announcing their model forthe structure of DNA.

In Franklin’s draft, it is deduced that the phos-phate groups of the backbone lie, as she had longthought, on the outside of the two co-axial helicalstrands whose geometrical configuration is specified,with the bases arranged on the inside. The twostrands are separated by 13 A (three-eighths of thehelix pitch in the axial direction). But the draftshows she had not yet grasped that the two chainsin B also ran antiparallel as in the A form. Her note-books show that for fitting the bases into the centreof a double helix, she had already formed the notionof the interchangeability of the two purine baseswith each other, and also of the two pyrimidines.She also knew the correct tautomeric forms of atleast three of the four bases, and was aware of Char-gaff’s base ratios. The step from interchangeabilityto the specific base-pairing postulated by Crick andWatson is a large one, but there is little doubt thatFranklin was poised to make it.

What would have happened if Watson and Crickhad not intervened with their great bursts ofinsight (Figure 27), and Franklin had been left toher own resources? It is a moot point whether shewas one and a half or two steps behind, and howlong it would have taken her to take them. Crickand I have discussed this several times. We agreeshe would have solved the structure, but theresults would have come out gradually, not as athunderbolt, in a short paper in Nature.

Figure 25. Excerpt from Frank-lin’s notebook on 24 February(Franklin papers). Annotations bythe author A.K. Franklin comesdown in favour of two chains forthe B form, making a connectionwith the A Form.

Figure 24. Excerpt from Frank-lin’s notebook on 23 February 1953(Franklin papers). Annotations bythe author A.K. Franklin begins toconsider a helical structure for theA form.


Pauling’s entry into the field

The story now moves back to Cambridge in early1953, when Crick and Watson re-entered the scene.News had reached them that Linus Pauling, thegreatest chemist of the day, had a structure forDNA and that a manuscript was on its way. Herean old rivalry asserted itself. In the early 1930sPauling and Bragg had been in competition aboutthe chemical basis of silicate structures. Then laterthere was the chagrin Bragg felt, as describedabove, at having missed Pauling’s a-helix.

Pauling’s manuscript arrived at the Cavendishin the last week of January, and it was immediatelyobvious he had made a crucial chemical mistake inpostulating a 3-chain structure with a central phos-phate-sugar backbone, and with the phosphatesunionized. It was chemically impossible, but nodoubt Pauling would return, or so Watson arguedto Bragg. (I doubt this—Pauling was a man withgreat insight, but not a magician, who could man-age without data).

Watson and Crick’s structure

The fact that Pauling was now in competitionmade for a race, and since the King’s groupseemed to be divided and making no progress,

Figure 26. Unpublished type-script dated 17 March 1953, whichis the precursor of Franklin andGosling’s paper in Nature 25 April1953 (Churchill College Archives;see Klug, Nature 1974, 248, p. 787).

Figure 27. Francis Crick and James Watson with theirmodel of the DNA double helix 1953 (Photo by Barring-ton Brown).


Bragg was persuaded to unleash Crick and Watsonfrom his earlier ban. Watson, two days earlier, hadvisited King’s to give a copy of the Pauling manu-script to Wilkins. It was then that Wilkins showedhim Franklin’s striking May 1952 X-ray picture ofthe B form, with its clear helical features. Thismade a profound impression on Watson, since onecould immediately count the number of layer linesleading to the 3.4 A meridional reflection. Herecounted this to Crick, along with the other par-ameters necessary to build a B form model: therepeat distance of 34 A, indicating ten units per

helical turn, a helix slope of 408, the diameter ofabout 20 A of the molecule, and they also remem-bered Franklin’s arguments for the backbonesbeing on the outside of the molecule and the baseson the inside.

The rest of the story is told in Watson’s vividaccount in his book, which revealed that Watsonand Crick had access to details of the informationin the MRC Subcommittee Report on the work atKings.. This was given to them in the secondweek of February by Max Perutz, a member ofthat Committee. The Report confirmed much that

Figure 28. Cylindrical lattice dia-grams of Watson and Crick’sdouble helix structure (right) andof the double helical structure Wat-son was aiming to build in earlyFebruary 1953 (left). The dots rep-resent individual nucleotide (orphosphate) positions projectedonto a cylindrical surface circum-scribing the helices, which is thenslit lengthways and unrolled flat.The oval symbols represent two-fold rotation axes of symmetry.

Figure 29. Analysis of Franklinand Gosling’s cylindrical Pattersonfunction map of the A form interms of a double helix. (Nature 25July 1953, 172, p. 156). The curvesdenote the “self Pattersons” of thetwo helical chains, separated byhalf the helical pitch. The fit isimproved if the “cross-Patterson”between the two chains is included(DLD Caspar, private communi-cation 1968).


they already knew, but the key fact was the spacegroup symmetry C2 of the A form. Franklin hadgiven this in her colloquium in November 1951,but Watson would not have understood it. Crickhad heard that the crystal was monoclinic, whichimplied a 2-fold axis of symmetry (a dyad), butthis could have been parallel or perpendicular tothe fibre axis. C2 required it to be perpendicular tothe fibre axis.

Watson had begun the building of two chainhelical models with the chains running in thesame direction (Figure 28, left). Each chain ofpitch 68 A would repeat after 20 nucleotides, butthe two chains were to be exactly half a helicalpitch apart, so that the structure would repeatafter 34 A in the axial direction. This would fit thebest estimate of the number of nucleotides per lat-tice point (16–24, deduced from Franklin’s densitymeasurements on the A form) which could bereconciled with the tenfold repeat. The chain hada helical rotation angle of 188( ¼ 3608/20) betweennucleotides, which brought successive sugarsclose together and was difficult to build. The C2symmetry, however, told Crick that there were

indeed two chains, that the chains ran in oppositedirections and that the helical repeat of ten unitsper turn referred to one chain of pitch 34 A, andso to each of the chains. Crick therefore changedthe rotation angle to 368 (Figure 28, right) and Wat-son found the chain easier to build. This was acritical step in getting the backbone structure right.

The formal account by Crick and Watson in theProceedings of the Royal Society (in 1954), whichdetails their cogent reasoning in arriving at thedouble helix, does not mention their knowledge ofthe crucial fact of C2 symmetry which they hadobtained from the MRC Report. It acknowledgesinformation received from Wilkins and Franklinonly in general terms: “We are most heavilyindebted in this respect to the King’s Collegegroup, and we wish to point out that without thisdata the formulation of our structure would havebeen most unlikely, if not impossible”. Presumablythere would have been some embarrassmentabout mentioning the source of their knowledge ofthe C2 symmetry. It would not have diminishedtheir achievement to have stated it.

The next step facing Watson and Crick was to fit

Figure 30. The sharpest, high res-olution A type diffraction patternobtained by Wilkins and the King’sgroup, post1953. (courtesy MauriceWilkins, in “Genesis of a Discov-ery”, ed S Chomet, 1993).


the bases stacked above each other into the middleof the double helix. The bases are linked by glyco-sidic bonds to the sugars of the backbones. Therewas room for two bases in each stack and Watsonhad been trying different ways of making suchpairs, connected by hydrogen bonds, initially pair-ing like with like, thus, adenine with adenine, andso on. In the last week of February, it was however,pointed out to Watson by Jerry Donohue, whoshared an office with him and Crick—anotherchance event—that he was using the incorrectchemical formulae (tautomeric forms) for the fourbases. When Watson changed these he found hecould fit in adenine-thymine as a pair, and alsoguanine-cytosine as a pair. The geometry of eachpair was almost identical!

Moreover each base-pair could fit either wayround between the two chains, A with T, and Twith A, and similarly for C:G and G:C. The glycosi-dic bonds were thus automatically related by theperpendicular dyod, thus fitting the C2 symmetry,although Watson had not made explicit use of thesymmetry in his model building.

Remarkably, this pairing also gave an expla-nation of the earlier finding by Erwin Chargaff

that the amount of adenine in any DNA sampleequalled that of thymine, and similarly for guanineand cytosine. Chargaff’s ratios thus automaticallyarose as a consequence of Watson’s base-pairingscheme. The structure of DNA was solved!

On 28th February 1953, Crick “winged” into theEagle pub, close to the Cavendish Laboratory,where lunch could be had for 1s 9d, and declaredto anyone who cared to listen that, in the Cavend-ish, Watson and he had discovered “the secret oflife”. Wilkins came to see their model in midMarch, and Franklin later at the end of the month.Her “instant acceptance amazed” Watson, butthen he did not know how far she had got towardsit, having heard only of her supposed “anti-heli-cal” stance.

There was agreement between King’s and Cam-bridge to publish separately, and three papersappeared on 25th April 1953, grouped togetherunder the overall title “Molecular Structure ofNucleic Acids”. Watson and Crick’s paper con-tained what appeared to be the famous throw-away sentence: “It has not escaped our notice thatthe specific [base] pairing we have postulatedimmediately suggests a possible copying mechanism

Figure 31. High resolution, sharp,X-ray diffraction pattern of a crys-talline B form, post 1953, obtainedfrom a lithium salt of DNA (Wilkinsin “Genesis of a Discovery”, ed. SChomet 1993). The outermost spotcorresponds to a spacing of 1.7 A,the second order of the spacing of3.4 A between the bases.


for the genetic material”. Crick explained later thatthey were not being coy, but there was a worry onWatson’s part that the structure might be wrong:when they sent the first draft of the paper toKing’s, they had not yet seen their papers and hadlittle idea of how strongly the King’s X-ray evi-dence supported their structure. After seeing itthey wrote their second Nature paper of May 30thentitled “Genetical Implications of the Structure ofDeoxyribonucleic Acid”) to spell out their postu-late for the copying mechanism in DNA replication(Figure 5). This paper also contains the first clearstatement on the genetic code: “The phosphate-sugar backbone of our model is completely regular,but any sequence of the pairs of bases can fit intothe [double-helical] structure. It follows that in along molecule many different permutations arepossible, and it therefore seems likely that the pre-cise sequence of bases is the code which carriesthe genetical information”

Proving the model

The first analytical demonstration of the generalcorrectness of the Watson–Crick model came inJuly 1953 from Franklin and Gosling (Figure 29).They showed that their Patterson function map ofthe A form could be fitted by a helical structurewith two chains.

The task of rigorously testing the model againstX-ray diffraction data required more accurateintensity data from better oriented fibre specimensand this was undertaken by Wilkins and the King’sCollege group including Herbert Wilson, Bob Lan-gridge and Watson Fuller. It took them aboutseven years to carry this out. They obtained muchimproved diffraction patterns from several differ-ent DNA sources (Figures 30 and 31), built higherresolution X-ray cameras, introduced computersto make the calculations and used new analyticmethods developed by Struther Arnot for refiningmodels to fit X-ray fibre diffraction.

During that time there were several objections bycrystallographers to the DNA model. These andother objections were finally answered by the rig-orous analysis at King’s, although other modelsappeared occasionally through the 60s and 70s.Indeed, it could be said that the formal crystallo-graphic proof of the double helix and the base-pairing did not come until 1979, when Drew andDickerson solved the structure of a dodecamericDNA oligonucleotide of defined sequence, byusing the totally objective heavy atom method(Proc. Nat. Acad. Sci. USA 78, 1981, 2178–83).

The reception of the double helix

It should be remembered that, in 1953, the X-raydiffraction crystallography of large biological mol-ecules was still in its infancy and regarded as anexotic pursuit; the first protein structures of myo-

globin and haemoglobin were not solved (at lowresolution) until 1957 and 1959, respectively.

The double helix model was well received bygeneticists and the phage group when Watsondescribed it at the Cold Spring Harbor meeting inthe summer of 1953, but there were doubts aboutthe correctness, and indeed relevance, of themodel on the part of biochemists, who, on thewhole, still thought of proteins as the geneticmaterial. The best biochemical proof that the struc-ture was correct eventually came from ArthurKornberg. If the “hypothetical” dyadic structure ofDNA with two antiparallel chains (Figure 3) werecorrect, then there must also be relationshipsbetween pairs of dinucleotides, further to Char-gaff’s rules for individual bases. Thus the numberof AG dinucleotides should equal the number ofCT dinucleotides, the number of TG equal to CA,and so on. Kornberg and his colleagues measuredthe frequencies of dinucleotides in a variety ofDNAs. The prediction was proved correct, in amost elegant way (Josse et al., J. Biol. Chem. 236,1961, 804–75).

Nevertheless, the structure of the double helix,as emphasized by Todd, was still only a discovery

Figure 32. The interpretation of the Messelson–Stahlexperiment, demonstrating semi-conservative replication(reproduced from L Stryer, Biochemistry, 4th Edition1995, Freeman).


in chemistry, and even if correct, the biologicalimplications for replication (“semi-conservationreplication”) as postulated by Crick and Watson,persuasive as they were, did not necessarily follow.The proof came in 1958 from “the most beautifulexperiment in biology” (the title of a recent book)by Messelson and Stahl (PNAS, 44, 1958, 671–675). This demonstrated unequivocally that thecomplementary strands of a DNA molecule separ-ate from one another, and that each strand thenserves as the template for the synthesis of a comp-lementary strand, duplicating its former partner,and so producing two DNA double helices (Figure32).

Biochemists and biologists generally also beganto understand that the Watson–Crick base-pairingallowed an infinite variety of irregular sequencesof the four bases of DNA to be accommodatedwithin the double helix, and so it was possible forDNA to act as carrier of genetic informationbased, on the four letter code A,G,C,T. In 1962, theNobel Prize for Physiology and Medicine was

awarded to Crick, Watson and Wilkins. RosalindFranklin had died in 1958, so the Nobel Committeewere spared the difficulty required by their sta-tutes of limiting the prize to a maximum of threepeople. The citation reads “for the discoveries con-cerning the molecular structure of nucleic acidsand its significance for information transfer in liv-ing material.” Note the word “information”, aterm that had never appeared in the writings ofbiochemists, who had been primarily concernedwith the transfer of energy in chemical reactions.Indeed, the citation looks forward to the geneticcode, research on which was well under way bythen.

The aftermath

As is usually the case with a fundamental dis-covery, the discovery of the DNA structure wasonly a beginning of a new epoch, the beginning of

Figure 33. Scheme of DNA replication. The DNA double helix (top) is cut at the replication fork by a topoisomeraseenzyme, and unwound by a helicase, the separated strands being coated with single strand DNA-binding protein(SSB). The leading strand (left) is copied into RNA (red) in straightforward way by the enzyme DNA Polymerase III(the “locomotive”). Since nucleic acids can be synthesized only in the 50 –30 direction, the lagging strand (right) is syn-thesized by an elaborate mechanism, using RNA intermediates, from short DNA sequences, by DNA polymerase I (the“sewing machine”). These are then linked covalently together by ligases. (Plate 14 in “DNA Replication, Second Edi-tion” (1992) by A. Kornberg & T. Baker, W. H. Freeman and Co., NY; courtesy of Arthur Kornberg).


molecular biology. Many major questions arose. Ican deal here only briefly with them.

I start with the problem of the replication ofDNA. The principle of semi-conservative replica-tion suggested itself to Crick and Watson directlyfrom the structure of the double helix and isstartlingly simple. But how does the helix actuallyunwind, and how does the sequence of each strandget copied? The implementation is startingly com-plex. Nucleic acids are only synthesized in onedirection (50 to 30): how then does the antiparallelstrand get copied? Again, the work of a generationof biochemists, notably Arthur Kornberg, hasshown that it takes dozens of protein complexes,each involving many proteins to accomplish this.They can be thought of as complex components ofseveral giant molecular machines (Figure 33),which synthesize the new DNA, check it for errors,and pass it on for further interactions which pack-age it in chromosomes.

There was a second major question. How doesthe information carried by the sequence of basesin a DNA molecule get finally transferred into thesequence of amino acids in a protein? The centraldogma was formulated by Watson as “DNA

makes RNA makes protein”, and by Crick as“sequence information can only pass from nucleicacid to protein and not in reverse”. This requiredthe genetic code to be worked out (Figure 34)which was largely accomplished by 1962. It hasfurther taken a generation of biochemists towork out the actual biochemical mechanismsinvolved in transcribing DNA into RNA. Theenzyme responsible is RNA Polymerase, ofwhich there are three varieties in eurkaryotes.The enzyme is another complex “molecularmachine” whose structure has recently beensolved by Roger Kornberg (Kornberg fils) (Figures35 and 36), and this enzyme acts only after a pre-initiation complex, involving dozens of other pro-teins, has been set up to recruit it to the gene tobe transcribed. The product RNA is then pro-cessed and passed as a messenger from the cellnucleus to the cytoplasm to ribosomes, the pro-tein factories which synthesize proteins ofdefined sequence. Here the message contained inthe sequence of the nucleic acid is translatedinto a sequence of amino acids according to thegenetic code (Figure 34), and also the polypeptidechain is assembled (Figure 37).

Figure 34. Transcription of theDNA double helix (represented bya two chain ladder) to make mes-senger RNA, the sequence ofwhich is then translated by the pro-tein synthesis machinery to make asequence of amino acids, by follow-ing the genetic code.


Epilogue

The discovery of the double helix and the elu-cidation of the genetic code launched the newsubjects of molecular genetics and, combinedwith biochemistry, the molecular biology of thegene. There also followed over the 50 yearswhat has been called the genetic revolution inbiotechnology but this did not stem directlyfrom the new knowledge. Rather it depended onthe development of tools for handling andmanipulating DNA. The key methodologicaladvances were Fred Sanger’s method of sequen-cing DNA, and recombinant DNA technologywhereby DNA molecules could be cut andpasted together in new combinations. Segmentsof DNA could be cloned and multiplied in bac-teria, and also used to express gene products inthem. To these must be added many otherpowerful methods, for example, the introductionof site specific mutations in DNA, and the poly-merase chain reaction which has replaced clon-ing for many purposes.

Then there have also been great advances inunderstanding the regulation of gene expression,that is, the switching of genes on and off in theright place at the right time by combinations ofprotein transcription factors, interacting with the

control regions of the gene. In higher organismsthe substrate, so to speak, for the expression isnot naked DNA, but chromatin in which DNAis packaged in nucleosomes, so there are com-plex mechanisms for making the control regionsaccessible to the transcription machinery. More-over since transcription factors working on agene are themselves the products of other genes,we really need to understand the networking ofgenes. This takes us on to the genome, to thehuman genome project and to the comparativegenomes of other organisms. There is muchmore to find out.

Acknowledgements

I thank Francis Crick, Maurice Wilkins & Ray-mond Gosling for helpful discussions. Thispaper is based on my lecture in Cambridge on25 April 2003 during the 50th Anniversary cele-brations of the Double Helix. That lecture wasan expanded version of an earlier one I gave inJanuary at Darwin College, Cambridge. I thankRichard Henderson for useful comments onboth.

Figure 35. Structure of RNA Poly-merase II, the central enzyme ofDNA transcription, viewed end onto show the cleft (at the top) forlocating the DNA double helix tobe transcribed (Cramer et al. (2001)Science 292, 1863; courtesy of RogerKornberg).


Figure 36. Cut-away side view of a complex of RNA Polymerase II, with a DNA double helix (blue) trapped in theact of transcription. The newly transcribed RNA (red) will exit from the top left-hand corner (Gnatt et al. (2001). Science,292, 1876; courtesy of Roger Kornberg).

Figure 37. The two subunits ofthe ribosome. The 30 S subunitwith the aid of tRNA, translates thesequence of the messenger RNAinto a sequence of amino acids,which are successively assembledinto a polypeptide chain. The twosubunits are linked physically andfunctionally by transfer RNAwhich, with one end, reads the gen-etic code on the 30 S subunit, and,at its other end, provides an acti-vated amino acid for peptide syn-thesis on the 50 S subunit.(Courtesy D. Brodersen andV. Ramakrishnan, MRC, Cam-bridge). From original diagrams:30 S: Wimberly, B. T., Brodersen,D. E., Clemons, W. M., Jr, Carter,A. P., Morgan-Warren, R. J., Vonr-hein, C., Hartsch, T. & Ramakrish-nan, V. (2000). Structure of the 30 Sribosomal subunit. Nature 407,

327–339. 50 S: Ban, N., Nissen, P., Hansen, J., Moore, P. B. & Steitz, T. A. (2000). The complete atomic structure of thelarge ribosomal subunit at 2.4 A resolution. Science 289, 905–920.


Further reading

Key papers

1. Watson, J. D. & Crick, F. H. C. A structure for deoxy-ribose nucleic acid. Nature, 171, 737 (April 25, 1953).

2. Wilkins, M. F. H., Stokes, A. R. & Wilson, H. R. Mol-ecular structure of deoxypentose nucleic acids.Nature, 171, 739 (April 25, 1953).

3. Franklin, R. E. & Gosling, R. G. Molecular configur-ation in sodium thymonucleate. Nature, 171, 742(April 25, 1953).

4. Watson, J. D. & Crick, F. H. C. Genetical implicationsof the structure of deoxyribonucleic acid. Nature(May 30, 1953).

5. Crick, F. H. C. & Watson, J. D. The complementarystructure of deoxyribonucleic acid. Proc. Roy. Soc.223, 80–90 (1954).

Other Relevant Papers in Historical Order

6. Watson, J. D. & Crick, F. H. C. The structure of DNA.Cold Spring Harbor Symp. Quant. Biol. 18, 123–131(1953).

7. Franklin, R. E. & Gosling, R. G. The structure ofsodium thymonucleate fibres. I. The influence ofwater content. Acta Crystallog., 6, 673 (1953). Sub-mitted March 6.

8. Franklin, R. E. & Gosling, R. G. Evidence for 2-chainhelix in crystalline structure of sodium deoxy-ribo-nucleate. Nature, 172, 156 (July 25, 1953).

9. Wilkins, M. H. F., Seeds, W. E., Stokes, A. R. &Wilson, H. R. Helical structure of crystalline deoxy-pentose nucleic acid. Nature, 172, 759 (October 24,1953).

10. Gosling, R. G. Thesis, University of London (1954).11. Langridge, R., Wilson, H. R., Hooper, C. W., Wilkins,

M. H. F. & Hamilton, L. D. The molecular configur-ation of deoxyribonucleic acid: X-ray diffractionanalysis. J. Mol. Biol. 2, 19 (1960).

12. Langridge, R., Marvin, D. A., Seeds, W. E., Wilson,

H. R., Hooper, C. W., Wilkins, M. H. F. & Hamilton,L. D. The molecular configuration of deoxyribonu-cleic acid: molecular models and their Fourier trans-forms. J. Mol. Biol. 2, 38–64 (1960).

Other historical references

13. Rosalind Franklin papers, in the Archives of Church-ill College, Cambridge.

14. Wilkins, M. H. F. Nobel Lecture in Le Prix Nobel en1962 (Stockholm).

15. Klug, A. Nature, 219, 808–810. see also pp. 883–844(1968).

16. Perutz, M. F., Wilkins, M. H. F. & Watson, J. D.Reproduction of MRC Report December 1952 anddiscussion on its Degree of Confidentiality. Science,164, 1537–1539 (1969).

17. Crick, F. H. C. Nature, 248, 766–769 (1974).18. Klug, A. Nature, 248, 787–788 (1974).19. Wilson, H. R. Trends Biochem. Sci. 13, 275–278 (1988)

& Wilson, H. R. Trends Biochem. Sci. 26, 334–337(2001).

Books

20. Watson, J. D. 1968. The Double Helix. AthenaeumPress, New York. Reprinted in Stent, G. S., ed. TheDouble Helix: Text Commentary, Reviews, OriginalPapers (Norton, W. W., New York, 1980).

21. Olby, R. C. 1974. The Path to the Double Helix, Mac-millan, London.

22. Judson, H. F. 1996. The Eighth Day of Creation: TheMakers of the Revolution in Biology, Cold SpringHarbor Laboratory Press, Cold Spring Harbor, NYExpanded Edition.

23. Chomet, S. ed. 1993. Genesis of a Discovery: DNAStructure, Newman Hemisphere, London (Accountsof the work at King’s College, London).

24. Sayre, A. Rosalind Franklin and DNA. (W.W. Norton,1975).

(Received 10 November 2003; accepted 10 November 2003)


Klug dna

Documents

chemical structure of

fibres of dna

formsof dna

dna moleculemight

proposed structure

dimensional structure

kings college bywilkins

xray diffractiondata