ROUGH DRAFT -

1.

A NOTE ON MOLECULAR CONFIGURATION, IN SODIUM W&OKUCLEATE

Rosalind E. Franklin and R. G. Gosling

17/3/53 o

Sodium thymonucleate fiblccs give two distinct types of

X-ray diagram. The first, corresponding to a crystalline
form obtained at about 75% relative humidity, has been des-

v

o At high humidities a

.A

cribed in detail elsewhere (.

) *

new structul'e, showing a lower degree of order appears, and
persists over a wide range of ambient humidity and water

content. The water content of the fibres, which are crys-

talline at lower humidities, may vary from about 50% to

several hundred per cent. of the dry weight in this 3tructure.
Other fibres which do not give crystalline structure at all,

show this lass ordered structure at much lower humidities.
The diagram of this structure, which we have called structure
By shows in striking manner the features characteristic of
helical structures ( 1 Although this cannot be taken
as proof that the structure is helical, other consfderations
make the existence of a helical structure highly probable.

Structure B is derived from' the crystalline structure A

when the sodium thymonucleate (NaDNA) fibres take up quantities
of water in excess of about 40% of their weight.
is accompanied by an increase of about 30% in the length of

the fibre, and by a substantial re-arrangement of the molecule.
It therefore seem reasonable to assume that in structure B

The change

the structural units of NaDNA (molecules or groups of molecules)
are relatively free from the influence of neighbouring groups,
each unit being shielded by a sheath of water. Each ur,it is
therefore free to take up its least-energy configuration
independently of' its neighbours and, in view of the nature of
the long-chain molecubes involved, it is highly likely that the

2.

general form will be helical (
pothesis of a helical structure, it is imlediately possLbls,
from the X-ray diagram of structure B, to make certain
deductions as to the nature and dimensions of tha helix.

1.

If we adg-pt the Ly-

From the angle between the straight lini?3 which C~A ba

drawn through the origin and the innermost maxima of tna Ist,
2nd, 3rd and 5th layer-lines, the diameter of tha halix can
be calculated. It is Pound to be about 20A. Since this
linear array of mexima is one of the strongest feat-iirss of
the diagram, we must conclude that a { cryst3llogra~~ic&l~~~
very important part of the moloculo lies on 8 helix Gf -zkL::
dlametor. This can only be the phosphata groups (OT, ~s?';iaps,
the phosphorus atoms) . Thus, if tbs structure is halicsl
we find that the phosphate groups lie on a helix of disaster
about 20A, and the sugar and base groups must accor5inSiy Sa

turned inwards towards the helical axis.

--. --- -

This is in agreement rkith the conclusion which '6'6 reached

previously by quite other reasoning (
whatever the structural unit, the phosphate groups ram2 '36 on

the outside.  Thero were two principal reasons for bellavlilg
this. The first derives from the work of Gulland and his
collaborators who showed that even in aqueous solution tha -CO
and -?GI2 groups of the bases are inaccessible and cannot 33
titrated, whereas the phosphate groups are fully acccsaible.

The second is our own obvervations on the way in which tka
structural units in the crystalline structilre A are r'loatcd
apart by an excess of water, tne process being a contirnuouri
one which leads to the fonmation first of a gel and ult'L~ria",ly

to a solution.
presumed to lie in *he phosphate groups; ( (CZH~O~ZPOZN~ acd
(C3H70)2PO$?a are highly hygroscopic) and tha simplest ex-
planation of the above proce3s is that these groups lie ofi tha
outside of the structurnl units. Furthermore tlm re&y ava"ils-
bility of the phosphate groups for interaction with p~ottlns
can also be explained this way.

1 , namely that,

The hygroscopic part of the molecule my ba

The above estimate of 20A diameter was baaed on the

as3umption of a single strand helix. That ia, the first
maximon the nth layer-line corresponds to the first zaxbnilm
in Jn(2Tii.R) .
u, r is the radius of the. helix and R the distance frors the

fibre-axis direction in reciprocal space.

Where Jn(U) is the nth order Bessei ?unction of

The strong meridional maximum at 3.4 A' lies accurately

on the 10th layer-line, From this new lines of maxim

eminate, as from the origin, crossing the origin series on 5Ls
5th layer-lim, corresponding to a Sg(u) for each series, 2sc-
firming that the second origin does lie on the 10t'n l&yeT-lim.

This then, indicates that there are 10 structural unit3 In om
. turn of the single-atrand helix. For a helix of diameter 20 A0

this gives a distance of 6A between neignbourlng units in oil8

molecule, which is a reasonable distance for t'ne P-P value in

NaDNA.   (this distance in a fully extended chain is 6.8 A').

If, instead of a single-strand helix we propose 2 equaiiy

spaced co-axial helical molecules, the first maxixu9 on i;La
nth layer-line correspondsto the f irst rnaxiinun in JZn(21,"r,?j.
Since our value of R is fixed and the first maximm In J~(x)
occurs at very nearly twice the value in x of ths first maxixaul
in Jl(x) (which gave us 2r h20A0) the value of 2r for a
2-strand helix must be 40 A'. The cross-section oi" the

helix would then be considerably greater thn that of tka
primitive cell in tb crystalline structure A, and this would
seem highly improbable. The same argument, with even nore
force, eliminates the possibility of 3 equally spaced co-axial

helica 1 molecules .

I

On the theory of a single-otrand helix, the sor5es of

equatorial maxima should correspond to the maxima of Jo(4S;~ sin e).
The maxima on our photograph do not, however, fit this function.
This is rather to be expected. For we know that tha helix so
Tar considered is only the most important member of a seTies of
co-axial helicies of different radii, the nonlj$issghatos pa~ta

of the molecule must lie on a series of co-axial  helicies

of smaller radii.  Following Crick, Cochran and Vand', the
structure factor on the nth layer-line for a series of co-

(here give definitions )

Simplifying this, for the ease of a whole number of residues
per turn of the helix, we readily obtain

It follows thaf

maxima on the layer-
th l&'rgest diameter,
of xj and xk smaller

From this it is evident that the innermost

lines will always be given by the helix of

containing

the terms -Zag Jn(xj) Jn(xk) for values
than the maximum being, in the region of the first maxima, very
small.
owing to the appearance of important neeative terms in the
expression for I.

Later maxima, however, may be obliterated or shifted

Thus, while we do not attempt'to offer a complete interpre-

5.

tation of the fibre-diagram of structure B, we may state the
following conclusions o The structure is probably helical.
The phosphate groups lie on the outside of the structural

unit, on a helix of diameter about 20 A. There are 10

phosphate groups per chain in one turn of the helix. The

structure does not contain more than one equivalent co-axial
chain, but the possibility of non-equivalent co-axial chains

-

is not eliminated.

The total absence of an inner maximum on the fourth

layer-line suggest8 `that if there are 2 non-equivalent co-
axial chains these are separated by2 of the fibi.2-axis
period, that is by %A in the fibre-axis direction.

8

9 1.34