[tha ae 29, wpe ‘y “ad. 62 8% 16.2 “AN: 30% A* G fe ee .W.= 930, QenaD tbr 1, Me ager kd, fod ia A R Rr readied . O4 wth aad, Se alee (.. bed, f Qe raf net tr / rf f 208 bd > Lt maarernk Owe Ba zal / “TS vate We = ee 1 Xn xy (q jo: 1B, ] \ 4 yal sp ~ Aerie | lA AA an 9.de ane ok Aye 15° Od Cer Prva t 2.55% £9915" 2.47 A ¢ tk T asus ten Qarenf ve 2.45. | an Ro t= 335- 2S ogoh | 40/7 J B* evs Trey ttch, t Ye hegrage ste ak ra Ore c ett 2 eo aXe a LL, 5. af. ound rrr we OW Lvs pack Oy x > be { Le en 2. att ¥ FR | x 4A orn= 2. . (Cop oR POG an ra Se meat) A or ae e(« conn tithing Pea. yew), fd) “- a Teme Olngra 0A age, Ral sds TRA nels O-O = aus B... 2.558 o: 3.16 2.49 te Lett | ™ + (Rebabl D4 a ane Lipwest, Nasneg ih Qeisch Bosak oe keine 7 pfs f)— aun Ro —m f5 ~ “e = 2.55 > “f) wo, xe Wer = = oe * 2 pg ese *°) - eed 722 bee os “é sae Ou “3 xe * °) teeta eae cou 2.7 eee ure soe wi ole ah 24 Oo Dud box veelak a! aiteh = Baca aOx34= = LR) So Se matin ms Uren. V0 Do tte (@) nd “The { \Dy Yr~p-2 ON Prati 665%, Toner Oates a o78 ,. tin Onatze 286 Soe aa hous, Dass g =-iriPen wisn? vA rowalt "3 raidco wer tone, Tring BRL War chews Neal, arene a 1) O00 tr | SEO Va tedeyn Sd x ek wire ie 20 sth Unser O Seep Pz 120° apres Apna) sae A Oe oie “etePny Me w%e | See ey ae st 9 = 78. ot A et gy pote Ae, as i! 2 s j \ yoy . pm Wife biseh Nue'g - ; ny } a4 “a fy Now and arovad ed « WP a QP OA Be aA a oi! “hee, Tareme 1° ar anwgllt FL pe nn pew aa. 4A 17). Wadhwa to (= 2.15 8 Translations ol @ze 3.40% P= g4rK At 8.5¢°54.2.35"2 0. oa ne a OV Vor 4 y Ve Hh ana 167 Tse len)” — : M. welats he A On Eph. PAR Lo 7 ee ae Sah ttn ot on Bak Req apacang daca clay an digg WT Peas a RDG Bases 4 0 Th. Po haa ward! area at 3. weap be Chemistry A Proposed Structure for the Nucleic Acids By linus Pauling and Robert B. Corey Gates and Crellin Laborstories of Chemistry,* C:lifornia Institute of Technology, Pasadena 4, Calif. (Communicated December 1952) The nucleic scids seem to be comperable in importance to the proteins, as constituents of living organisms. There is evidence that they are involved in the processes of cel] division and erowth, emé thet Uk they participate in the transmission of hereditary charecters he men B_ tembe important constituents of viruses, se-weld—ac-of-suetarie, 6 n understanding of the molecular structure of the nucleic acids should xm witiz be of value in the effort to understand the fimd:nentel bkoakopinek BEOKEXSeHxX phenomena of himkuxzyx life. Only recently hasyconplete informetion been gathered about the chemi- cal nature of nucleic acids. Yheyxmamsictxek The nucleic acids are giant molecules, composed of complex units. Each unit consists of a phosphate ion, HPO; , a sugar (ribose in the ribonucleic acids, deoxyribose in the ee deoxyribonucleic acids), and a purine or pyrimidine side chain aprons = mimumeine (adenine, guanine, thymine, cytosine, wractt). The purine or pyrimidine is attached to carbon aton Y of the sugar. C tinoyer the investigations of Todd and his collaborators$* ‘ae Good evi-= et dence been obtained as to the nature of the linkage between the sugar snd the phosphateg it seems likely thet the phosvhete ester links involve carbon atoms 3/ and 5 ‘of the ribose or deoxyribose. X-rey photographs have been made of sodium thymonucleate and other preparations of nucleic acids by Astbury and Bell” and, more recently, by 3, Some information about the nature of the struc- tures has been obtained from these photographs, but it has not been found possible to derive detailed structures from the x-ray datae We have now formulated a promising structure, by making use of the general principles of molecular structure and the availeble information about the nucleic acidse Th structure is not a vegue one, but is precisely predicted; atomic coordinates for the principal etoms are given oe Wad \ FRA ve ee Fuk prud, Nercrleedd Keaton in thresfetteningaaes. ‘The structure frezinsh, for some of the features of ally VIRAL the x-ray photographs; JR intensity celevlations herve ae been made, -3- amhaliovedy end the structure cannot be considered to have been proved to be correct. The formulation of the structure. - The most important configuration of polypeptide chains in proteins is the a helixe“ In this structure the aminowacid residues are equivalent (except for differences in the side chains); there is only one type of relation between a residue and neighboring residues, one operation which converts a residue into a following residue. Through rotation= the continued application of this operation, a xmkakmxryxtranslation, the a helix is built wp. It seems not unlikely that a single general operation asymmetric is also involved in the construction of nucleic acids fron their/fundamental units. The general operation involved would be a rotation-reflection, and its application would lead to a helical structure. We assume, 2ccordingly, that the structure to be formulated is s helix. The giant molecule would trus be cylindrical, with approximately circular cross section. Some evidence in support of this assumption is provided by the electron micrographs of preparations of sodium thymonucleate described by Williams .° The preparation seen in the shadowed electron micrographs is clesrly fibrous/ in nature. The small fibrils or molecules seem to be -/= Gar A . circular in cross section, thet their diameter is apparently constant, there is no evidence that the molecules are ribbon-like. The diameter as estimated from the length of the shadow tm has been estimated at 15 or 20 Ae The x=ray photographs of sodium thymonucleate show a strong equatorial reflection at 16.2 4. If it is assumed that this is due to a hexagonal pack- ing of cylindrical molecules, the diameter of the molecules is 1867 4. From the average residue weight of sodium thymonucleate, about 330, and the density, =_2 about 1662 g cm “, we calculate that the volume per residue is 338 a, The cross-sectional area ts per residue is 303 a’; accordingly the length ver residue along the fiber axis is about 1.12 A. very The x-ray photographs show a/strong meridional reflection, with spacing about fadhbck 3.40 A. This reflection corresponds to a distance equal to along the fiber axis/three times the distance per residue. Accordingly, the reflection is to be attributed to three residues. If the molecule of a nucleic acid is 2 single helix, the reflection at 3.4 4 would heve to be attributed to a regularity in the purine-pyrimicine Sequence — that is, to a regular secuence of nucleotides, involving repetition structural of a/unit of three nucleotides. It seems unlikely that the nucleotides re- peat in this regular way; it is likely instead thet the nucleic acids, like the proteins (insulin), involve a less regular secuence of the funda- mental units. The alternative explanation of the strong 3e4-4 meridional reflection is that the cylincricel molecule is formed of three cheins, viich are coiled about one another. ‘The structure describec below is a three- chein structure, exch chain being a helix with fundemental trenslation ecual Each of the three helical chains is tightly coiled, with a little OTC Qgeutbibibekesss than three residues per turn of the helix. The pitch > of the helix representing a single chain is Sperevivmrtekp Je A. ‘the three ta chains interpenetrate, in such a way that the Pee@ of the triple helix is 434% abowls SEH i. The first question to be answered is that as to the nature of the core of the three-chain helical molecule - the part of the molecule closest to the axise It is important for stability of the molecule that atoms be well paced together, incuemmsxikkeky and the problem of packing atoms together is a more cifficult one to solve in the neighborhood of the axis than mf at a distance away from the exis, where there is a larger distance between -Ce an atom and the equivalent atom in the next units (in exenle of a helical structure which seems to im satisfy all of the structural require- ments except that of close packing of atoms in the region near the helical axis is the 5.2-residue helix (the ¥ helix) of volypeptide chains. This structureg seems not to be represented in proteins, whereas the May packed in a similar qa helix, in which the atoms are/satisf: ctorily close manner about the axis, is an important structure.) There are three possibilities as to the comoosition of the core; it may consist of the morax purine-pyrimi dine sd) A ye? Tra frond aa) Lar groups, the sugar, oF the phos: . cause of their varied nature, At tT» improbable that the purine-pyrimicdine groups coud be packed slong the axis of the helix in such a way that suitable bonds could be formed between the sugar and the onl Ret os pusxibitixyx choice is accordingly owe @liminated. It is yunlikely that the sugar groups axe constitute the core of the molecule’, The shape of the ribose molecule and the deoxyribose mole~ cule is such that close packing of these molecules alone a helical axis is ond wx aekiafoting wo 7, pacl—_ ah. Vas difficult, 4n exemple m® that shows the difficulty of achieving close ' ~ packing is vrovided the polysaccharide starch, which forms helixes with s fay a & "e 3 a hole tmm along the axis, into which iodine molecules can fite We conclude that the core of the molecule is probably formed of the phospho@&c Al gTouvSe & close-packed core of phosphoric acid residues, HPO,” , can easily be formed. At each level along the fiber axis there are three phosphate groups. These can be packed together in the way shown in Figure 1. Two oxy- gen atoms of each tetrahedral phosphate group form an octahecron, the trigonal which is the ficuer axis of the exis of marinxtwekrahedxabocphosphakex three-chain helical molecule. A similer complex of three phosphate tetrahedra can be superimposed on this one, with yt As change in azimuthel orientation. ‘The neizhoorhood of the axis of the molecule is then filled with oxygen atoms, arranged in groups of three, which change their azimathal orientetion by about 60° from layer to Lowe layer, in such a way as to produce closest packing of these atoms. The altitude of a phosphate ion (RO-| tees} is 17f Ae If the same distance were preserved between the next oxygen layers, the basal-plene distance along the fiber axis would be sud n This value is xexxx e¢toseto the spacing observed for the principal meridional reflection, suggmxbiagx It is to be exvected that the outer oxygen atoms of the comolex of three phos= -6- phate grouns would be attached to the ribose or deoxyribose, and that the hydrogen atom of the HPO,” residues would be attached to an inner oxygen aton, and presumably would be involved in hydrogen-bond formation with an- other of the inner oxygen atoms. The lenrth of the O-H...0 bond should be close to that observed in KH2P0,, 2.55 4. The angle P-O-H should be epproxi- mately the tetrahedral anglee It is found that the snacing 3.4 4 is not preserved, with this bond angle, if the hydrogen bonds are formed between one (HPO, ) 3 group and the groun above or below. Accordingly we assume that hydrogen bonds are formed between the oxygen atoms of the phosphate groups in zk the same basal plane, as indicated in Figure le If the bond angle P-O-H is assumed to be the tetrahedral angle, and the hydrogen bonds O-Hee*O are assumed to be linear, the phosphate groups must be rotated by 667°, in such a direction as to bring the plane of the inner oxygen atoms closer to the plane of the vhosphorus atoms. The g parameter of the inner oxygen atoms then becomes + 0e76 A, with that of P equal to 0.00 A. The g parameter of the outer oxygen atoms is + 0.96 A. -J= The radius of the inner oxygen atoms (the distance from the axis of the molecule) is found to be 2.11 4, assuming the values given above for the P-O and O-Hese0 distances. The parameters of the phosphorus atom and the outer oxygen atoms are easily calculated, and are given in Table le If the oxygen atoms in the next layer are placed at ecuel distances from those in the first layer, it is found that the group of three tetra- hedre is to be rotated through 6°, a while being translated by 3.40° along the zg axise The oxygen-oxygen contact distances are 2.45 4 (in the phos- phate tetrahedron), 2.55 & (O-Hees0 distance), 3045 4 (in the basal plene), and 2.74 4 (diagonal distance, between (EPO,), groups). It is found that a ribose residue mey be bridged ecross between the uoner oxygen atom of a tetrahecron and the lower oxygen atom of the tetrahedron above it, and rotated by approximately 120° (114° or 126°) in azimuth. The bridging may be achieved for either the right-handed screw arrengement of phosphate tetrahedra, shown in Figure 1, or the left-handed screw, the mirror image the of this. However, the right-handed screw secms to be/better, in several respects. In order to form ester linkages with carbon atom 2 and cerbon -8 atom 5 of the ribose residue, with the furanose-ring configuration, the olane of the 5=membered ribose ring must be pleced nearly at right angles to the basal plane (perpendicular to the axis of the m nucleic acid molecule), if the left-handed configuration is used for the phosphate complexe There then occurs steric hindrance between the ribose residue and the simi- lar residue almost directly above it - the rotation by 6° corresnonds to a lateral translation, at the radius (about 6 A) of the center of the ribose ring, of only about 1 A, which is not enough to permit the atoms of the two residues to clear one anothere For the right-handed configuration of the phosphate complex the plane of the ribose ring is at about 45° with the basal plane, and satisfactory packing of the suger residues is achieved. Also, the angle between the @xixandamnx C1-N axis, where N is the nitrogen atom of the purine or pyrimidine group, and the basal yene is about left <5° for the xkgkt-hended phosphate complex, and about 10° for the right- handed complex. The nucleic acids are observed to have strong negative birefringence. This anisotropy in apkixat index of refraction is to be ate tributed Taxgeinocteom almost entirely to the purine and pyrimidine planes, and it provides strong evidence that the pianes of these conjugated systems Je structure involving are nearly parallel to the basal plane of the molecule. The/right—handed phosphate complexes accordingly provides a more satisfactory explenation of the bir: fringence than does the other structures Coordinates of the atoms of the ribose murkewsx residue and of the nitrogen atom of the purine or pyrimidine group are given in Table le These coordinates are subject to gre:ter uncertainty than those for the phosphete groupse The way in which the ribose residue bridges the region between one XOXO KERNEL RE LCE LMI phosphate group and the next in the nucleic aecic chain is shown in Figure 26 Description of the structuree - In the vroposed structure each nucleic acid chain forms a tightly coiled helix, with approximately three ribose-phoa- phate residues per turn of the helix. The lead of the helix (the distance along the fiber axis from one ‘knrmurfixkkexketix position on the chain to the corresnonding vosition on the same chain after one complete turn) is ap= identical proximately 10 4. Three/chains are intertwined, z A single chain is re= he - presented in Figure 3. In the complete molecule three idential chains are Aw intertwined, as shown in Figure 4, to form a closely packed three-chain helical molecule. The three chains are attached to one another by leteral ~LO- hydrogen bonds between the inner oxygens of the phosphate groups. The diameter of the three-chain molecule, taking into consideration the size of the purine-pyrimidine groups, is about 20 A. If the oxygen atoms in one phosphate complex are equidistant from basic neighboring oxygen atoms in the next complex, the xmtakiamaz angle of rota- 4n individual tion of the helix differs from 120° by 6°. ®km/chain accordingly has ean identity distance of approximately 60 times the axial length per residue, 3e4 4¢ The identity distance of the three—chain molecule is predicted to be one third as great, about 20 times 3.4 4 = 68 4. Bee The identity distance cannot be predicted very accurately, however, because a consider- able change can be made without causing the oxygen-oxygen contact distances to be unsatisfactory. If the rotation wmmmd@@ differed from 120° by 12°, rather than by 6°, the oxygen-oxygen contact distances would be 2069 A and 2-77 A, respectively. These values are acceptable, and accordingly the identity distance might be that corresponding to a 12° rotation, which is 10 times 3.4 4 = 34 A. The x-ray photographs indicate an identity distance along the fiber axis of approximately 50 A. -ll- We plan to make a detailed comparison of intensities and other features of x-ray photogrpphs of nucleic acid prepzrations, and the pmertksx kKimmex calculated values for the proposed structure. It should be véssible to eliminate the structure, or to obtain further support for it. This investigation was aided by grants from The National Found