907

Acta Cryst. (1963). 16, 907

T h e Crystal and Molecular Structure of a Hydrogen-Bonded Complex Between 1-Methylthymine and 9-Methyladenine*

BY KA_RST HOOGSTEEN

Gates and Crellin Laboratories of Chemistry, California Institute of Technology, Pasadena, California, U.S.A.

(Received 25 September 1962)

Crystals of a 1:1 hydrogen-bonded complex between 1-methylthymine and 9-methyladenine can be grown from an aqueous solution containing equimolecular quantities of the two compounds. The crystals are monoclinic, with a--8-304, b = 6.552, c = 12.837 A, and fl--106 ° 50'. The space group is P21/m, with two base-pair complexes in the unit cell. The structure was refined with three-dimen- sional data taken with copper radiation. The positional coordinates and anisotropic temperature factors of the heavy atoms were obtained by least-squares analyses. The hydrogen atoms, except those of two methyl groups, were located from a three-dimensional difference Fourier synthesis. The 1-methylthymine and 9-methyladenine molecules form a planar base pair lying in a mirror plane and are connected to each other by two nearly linear hydrogen bonds, from the NH 2 group of 9-methyladenine to 0(9) of 1-methylthymine (2.846 A) and from N(3) of 1-methylthymine to N(7) of 9-methyladenine (2.924 A).

This structure differs from the adenine-thymine pairing proposed by Watson & Crick, where N(3) of thymine is hydrogen bonded to N(i) of adenine. The distance between the methyl group at N(1) of 1-methylthymine and the one at N(9) of 9-methyladenine is 8-645 A, whereas this distance is 11.1 /~ in the pairing proposed by Watson & Crick.

I n t r o d u c t i o n

As p a r t of a p rogram of research on the s t ruc ture of nucleic acids, a t t e m p t s have been made to grow crystals containing hydrogen-bonded pairs of purine and pyrimidine derivatives. A pre l iminary report on the s t ruc ture of a 1:1 complex between 1-methyl- thymine and 9-methyladenine (I) has a l ready been given (Hoogsteen, 1959). The manner in which these

H(4) / • H(3)--N(10') N(I') H(5)

H(8, 9) ~ / ~ / / .......... C(6') C(2') H(7)--C(IO) 0(9) II I

~ C(5') N(3') C (5)--C(4) / ~C(~')

H(1)--C(6) N(3)--H(2)-.--N(7') I / ~ / N ( 9 ' ) N(1)--C(2)

/ ~ / C ( 8 ' ) ~ H(10)--C(7) 0(8) H(6) C( i i ' ) - -H( i4 , i5) I-I( 1 l, i2) / H(13)

(I)

two molecules are hydrogen-bonded to each other was established with the use of hOl and hll intensities. They form a p lanar base pair wi th N(3) of 1-methyl- thymine connected to N(7) of 9-methyladenine, and the C ( 4 ) = O earbonyl group of 1-methyl thymine

* Contribution No. 2888 from the Gates and Crellin Laboratories of Chemistry. This investigation was supported, in part, by two grants to the California Institute of Tech- nology, a grant CVRE 121 from the National Foundation, and a grant :No. G-9467 from the :National Science Founda- tion.

hydrogen-bonded to the amino group of 9-methyl- adenine. This paper describes the ref inement of the approx imate s t ruc ture wi th the use of three-dimen- sional da t a and discusses the conclusions t h a t can be d rawn from the geomet ry of the base pa i r as i t occurs in the crystals .

E x p e r i m e n t a l

Equimolecular quant i t ies of 1-methyl thymine and 9-methyladenine were dissolved in hot water . Slow cooling and subsequent evapora t ion of this solution to dryness a t room t empera tu re gave good crystals of the 1:1 complex in the form of well developed needles, e longated in the direction of the b axis. A needle bounded by the forms {100}, {O01}, and {010}, wi th a cross section of 0.2 ×0.2 ram, was moun ted with the ro ta t ion axis paral lel to the b axis. This crys ta l was used in the p repara t ion of mult iple-fi lm equi-inclination Weissenberg photographs of layers hO1 th rough h5l. By cleaving a needle paral le l to (010), a f r agment wi th approx imate ly the same cross section as the first crystal was obtained and mounted paral lel to the a axis. This crystal was used to collect the intensities of layers Okl th rough 7kl. Ni-fi l tered C u K a radia t ion was used. The intensities were es t imated visually wi th the aid of a calibration strip. Af ter correction for the Lorentz and polarizat ion factors, the two sets of d a t a were correlated to the same relat ive scale. In all, 1361 reflections were measured with an in tens i ty different f rom zero, of which 856 were obta ined about both the a and the b axes. These reflections were used to obtain an exper imenta l

908 COMPLEX BETWEEI~ 1 -METHYLTHYMIIqE AND 9 - M E T H Y L A D E N I N E

estimate of the standard deviation in the observed structure factors, ¢(Fo). The values of

A F = ½(IF~I- lEVI)

where Fo ~ and Fbo are the structure factors measured along the a and b axes, were averaged over 6 regions of similar Fo values and plotted against the average value of Fo for that region. These points were then fitted by the least-squares method to the equation AF=A+BFo+CF2o, giving A=0.27, B=0.012, C= 0-0012. The coefficients A and B were used in the final stages of the refinement to calculate the weights of the reflections.

Accurate cell dimensions were determined from zero-layer Weissenberg photographs taken around the a and b axes. A Weissenberg camera with the Straumanis method of film loading and a radius of 10 cm was used. Observed values of sin 9' 0 for 60kl and 12 hO1 reflections, occurring in the back reflection region where the splitting of the c¢1 and ~ reflections was complete, were used as input to a least-squares treatment; the cell dimensions are:

a = 8.304 _ 0.002, b = 6.552 +_ 0.002, c = 12.837 _+ 0-003 A; fl = 106 ° 50' _ 3';

Dx = 1.438, D,n = 1-433 g.cm -3

for 2 Cu Kc¢1 = 1.54051 A. The standard deviations were derived from the least-

Squares calculation in the usual way.

Refinement of the s t ruc ture

In the preliminary paper (Hoogsteen, 1959) an account was given of the structure determination and the partial refinement with the hO1 and hll data until the disagreement index was 0-165. Two additional least- squares analyses were calculated in which the posi- tional parameters x and z together with the individual isotropic temperature factors were adjusted. The disagreement index decreased to 0" 145. The refinement was then continued with three-dimensional data on a Burroughs 220 computer.

The monoclinic structure-factor and least-squares program allows for the refinement of the positional parameters, six anisotropic temperature factors and an individual scale factor for each atom. I t determines the minimum value of the expression Xw(IF~I-IF~]) ~" and uses the scattering factors of Berghuis, Haanappel, Potters, Loopstra, MacGillavry & Veenendaal (1955) for C, N, O, and of McWeeny (1951) for H. The normal equations are solved with neglect of all interatomic cross terms. For each atom only cross terms between the x and z parameters and between the six anisotropic temperature factors together with a scale factor are collected. The space group of the crystals had been found to be P21 or P21/m (Hoogsteen, 1959). If it is assumed to be P2dm, all carbon, nitrogen, and oxygen atoms of the base pairs must lie in the mirror planes

at y = ¼ and y = ~. The positional parameters x and z, the anisotropic temperature factors, c~, fl, y, and e and one scale factor remain to be refined. The scale factor was adjusted by taking the average value of the indicated corrections to the individual atomic scale factors. The weights were calculated according to the Hughes (1941) weighting scheme-

~/w = 1/F~ if Fo > 4Fmin,

~/w = 1/4FoFmi, if Fo <_ 4Fmin.

These weights do not accurately reflect the un- certainties of F~ values, but they were considered to be adequate at this stage of refinement. Accidental absences were included in the refinement calculations only if their calculated values exceeded the minimum observable ones. After five refinement cycles the disagreement factor decreased to 0-125.

At this point it was decided to at tempt the resolu- tion of an ambiguity resulting from the configuration of the 1-methylthymine molecule. With only minor changes in the atomic positions, the crystal structure could accommodate base pairs (II), of which 0(9)

H(4) /

• H ( 3 ) - - N ( 1 0 ' ) N ( l ' ) H ( 5 ) . . . . ~ / ~ /

.... C(6") C(2") C(7) O(8) II !

~ C(5') N(3") N(1)--C(2) / ~C(4')

H(1)--C(6) N(3)--H(2) -.N(7') I C(5)--C(4) \ /

/ ~ /C(8') C(ll') C(10) 0(9) H(6)

(II)

instead of 0(8) is hydrogen bonded to H(4) and 0(8) instead of 0(9) is hydrogen bonded to H(3). The principal difference between this structure and the one assumed until so far is an interchange of the atoms N(1) and C(5) of the 1-methylthymine molecule.

The same type of structural ambiguity is possible in the structure of 1-methylthymine (Hoogsteen, 1963), but the positions of N(1) and C(5) could be identified at a similar stage of refinement by means of a difference Fourier synthesis, and subsequent calculations gave no indications of disorder. A three-dimensional dif- ference Fourier map of the present structure was therefore calculated. In this difference m~p the electron densities at the positions of N(1) and C(5) were not significantly different from those at other atomic positions, indicating that the configuration initially chosen (I) was the correct one. The positions of all hydrogen atoms directly attached to the ring systems of the adenine and thymine molecules were also indicated in this map by distinct maxima. One hydrogen atom of each of the three methyl groups could also be located in the mirror plane, and it was therefore assumed that the orientation of the methyl groups obeyed the symmetry restriction of the mirror

: K A R S T H O O G S T E E N 909

Tab le 1. Fina l coordinates and calculated estimated standard deviations of the P2~/m least-squares refinements, and anisotropic temperature factors

The coordinates are listed in fractions of the unit-cell axes; y---- ¼

X O'(X) (A) z ty(z) (A) 1040¢ (7(0¢) 104fl (~(fl) 1047 a(7) 104e a(e) :N(1) 0.7485 0.003 0.4370 0.003 143 (6) 285 (11) 77 (3) 79 (6) C(2) 0.7598 0.004 0.5461 0.004 125 (7) 275 (12) 70 (3) 35 (6) N(3) 0.6078 0.003 0.5693 0.003 124 (5) 284 (10) 52 (2) 33 (5) C(4) 0.4506 0.004 0.4847 0.004 129 (6) 261 (11) 65 (3) 35 (6) C(5) 0.4481 0.004 0.3835 0.004 147 (7) 290 (13) 64 (3) 38 (6) C(6) 0.5936 0.004 0.3591 0.004 164 (7) 293 (13) 60 (3) 55 (7) C(7) 0-9024 0.004 0.4043 0.005 152 (8) 399 (17) 101 (4) 116 (9) O(8) 0"8905 0"003 0.6184 0"003 127 (5) 444 (12) 89 (2) 18 (5) 0(9) 0.3234 0"003 0.5265 0.002 138 (5) 408 (11) 59 (2) 53 (5) C(10) 0.2781 0.004 0.2991 0.004 167 (8) 394 (15) 52 (3) 17 (6)

N(I') 0.2659 0.003 0.9123 0.003 171 (7) 299 (11) 69 (3) 76 (6) C(2') 0.3687 0.004 1-0164 0.004 191 (8) 318 (13) 63 (3) 73 (7) N(3') 0.5324 0.004 1.0542 0.003 181 (7) 312 (11) 58 (2) 45 (6) C(4') 0.6012 0.004 0.9712 0.003 157 (7) 204 (10) 62 (3) 38 (6) C(5') 0.5184 0.004 0.8618 0.004 155 (7) 259 (12) 61 (3) 38 (6) C(6') 0-3416 0.004 0.8319 0.004 162 (7) 236 ( l l ) 63 (3) 51 (6) N(7') 0.6310 0.003 0.8010 0.003 145 (6) 335 (12) 63 (2) 56 (6) C(8') 0.7807 0.004 0-8742 0.004 155 (7) 314 (13) 75 (3) 49 (7) N(9') 0.7714 0.003 0.9785 0.003 146 (6) 279 (11) 67 (2) 16 (5) N(10') 0.2461 0.003 0.7285 0.003 128 (6) 465 (14) 61 (2) 25 (3) C(ll ' ) 0.9088 0.004 1.0789 0.004 160 (8) 423 (17) 74 (3) --21 (8)

p lane. The coord ina tes of t he h y d r o g e n s l y i n g out of the mi r ro r p lane were t h e n ca lcu la ted , a n d these pos i t ions were found to cor respond to e x t e n d e d a n d weak m a x i m a in t he di f ference map . The coord ina tes of t he h y d r o g e n s t h u s o b t a i n e d were t h e n i n s e r t e d in t he s u b s e q u e n t l eas t - squa res r e f i n e m e n t s ; t h e y were ass igned one i so t rop ic t e m p e r a t u r e fac tor w i t h B = 4.0 .~2.

The r e f i n e m e n t was comple ted b y 8 more leas t - squares cycles in which the we igh t s were ca lcu la t ed accord ing to t he express ion

1

~/w = 0.50 + 0.54Fo + 0-024Fo ~ "

The coeff icients of Fo a n d Fo ~ were de r ived f rom A a n d B of t he e x p e r i m e n t a l l y d e t e r m i n e d e q u a t i o n for a(Fo). The c o n s t a n t 0.50 was a d d e d to decrease the we igh t s of t he weak a n d u n o b s e r v e d ref lect ions . The ca lcu la t ed correct ions to t h e i n p u t p a r a m e t e r s of t he h y d r o g e n a toms were also appl ied . The ave rage cor rec t ion for t he pos i t iona l p a r a m e t e r s of t he h e a v y a toms for t he f ina l ca lcu la t ion was 0.0004 A w i t h a m a x i m u m va lue of 0.001 X. :None of t he correc t ions of t he o the r p a r a m e t e r s was la rger t h a n 0.2 t imes t he i r s t a n d a r d dev ia t ions , a n d r e f i n e m e n t based on the c e n t r o s y m m e t r i c space g roup was t e r m i n a t e d a t t h i s po in t . The d i s a g r e e m e n t i n d e x R was 0.081, w i t h exc lus ion of t he unobse rved ref lect ions . The f ina l pos i t i ona l p a r a m e t e r s x a n d z, t o g e t h e r w i t h the an i so t rop ic t e m p e r a t u r e fac tors a , fl, ~, a n d ~* for t h e carbon, n i t rogen , a n d o x y g e n a t o m s o b t a i n e d in t h i s way , are l i s t ed in Tab le 1.

D u r i n g the r e f i n e m e n t the i n t e r a t o m i c d i s t ances

* These temperature parameters are the coefficients of the expression T = exp - - ( ~h ~ + flk ~" + yl 2 + 6hk + ehl + ~Tkl).

a n d bond angles i nvo lv ing the h y d r o g e n a t o m s were ca lcu la ted as a check on t h e i r pos i t i ona l coordinates .

z 0

/ ~ 7 ( ° ,-- ~ , " ~ I . . . . . / ' \

/ / (-~x_~' /-".,~ t ( ) )

/. 2" n) :J <=, ( ,,'h. l~ " , k ' ~ J I I ~ I ~ ~--~ / / " ',~ . t'-'~

/--K" \ - / ) t f",ffk"~ I ,'A, r \ ~ - : I

l) I i r-,t ,-.. (~/- "Y i J /,'

7"7"-o',.-" ' U, ~ ~, .-.<-~"~t(OJ..m t - ~ - /U'.: "° ,..-:.:t .,,_J> {

t # 3

1 ( - , - J ,-, /

Fig. 1. The section parallel to (010) at y = ¼ of the final dif- ference-Fourier synthesis. Contours are drawn at intervals of 0.2 e.A -3. Broken and dotted lines indicate the zero and negative contours.

A l t h o u g h the d i s t ances were no t un reasonab le , i t became clear t h a t t he h y d r o g e n a t o m s l y ing above a n d below the mi r ro r p l ane were m o v i n g t o w a r d s t he mi r ro r p lane . The angles wh ich t h e ca rbon a t o m s s u b t e n d to those h y d r o g e n a toms were as a r e su l t too smal l , a b o u t 80 ° for t he l a s t r e f i n e m e n t ca lcu la t ion . A f ina l d i f fe rence-Four ie r s y n t h e s i s was there fore ca l cu la t ed based u p o n the p a r a m e t e r s of Tab le 1; the con t r i bu t i ons of t he h y d r o g e n a toms to t he s t ruc- t u r e fac tors were omi t t ed . The sec t ion a t y = ¼ , pa ra l l e l to (010), i.e. t h r o u g h the mi r ro r p lane , is shown in Fig. 1. The h y d r o g e n a toms t h a t are d i r e c t l y a t t a c h e d to t h e r ing s y s t e m s of t h e t h y m i n e a n d aden ine molecules a l l show up as m a x i m a w i t h a he igh t of a b o u t 0.5 e./~ -3. The i r coord ina tes were d e t e r m i n e d f rom th i s m a p b y the i n t e r p o l a t i o n m e t h o d of B o o t h (1948); t h e y are l i s t ed in Tab l e 2. I n Fig. 2

910 C O M P L E X B E T W E E N 1 - M E T t t Y L T H Y M I N E A N D 9 - M E T H Y L A D E N I N E

Table 2. Coordinates of hydrogen atoms determined from the final difference-Fourier analysis with space group

P21/m The temperature parameters ~ere derived from

the final least-squares calculation

x y z B H(1) 0.596 0.250 0.642 3.6 A 2 II(2) 0.595 0.250 0.282 2.0 H(3) 0.275 0.250 0.652 3.7 H(4) 0.137 0.250 0.711 2-9 tI(5) 0.287 0.250 1.061 2.9 II(6) 0.881 0-250 0.857 3.4 tI(13) 0.018 0.250 0.060 3.7 H(14) 0.884 0.127 0.122 5.7

H12

I-I1

(CH3)7 (a)

H8

H9

(CH3)1o (b)

14

H13

(CH3)~11

(c)

Fig. 2. Sections through the electron density associated with the hydrogen atoms of the three methyl groups. Contours are at intervals of 0.2 e./k -3, beginning with the 0.2 e./~ -3 contour.

are shown sections through the planes containing the hydrogen a toms of the three methyl groups a t C(7), C(10), and C(1 l ') . Al though the three m a x i m a around C( l l ' ) in Fig. 2(c) are low (about 0.4 e./~ -3) their coordinates were de termined and accepted as the coordinates of the hydrogen a toms of C( l l ' ) . They are also listed in Table 2. The hydrogen coordinates obtained so far did not differ significantly from those derived from the least-squares analysis, and the t empera tu re pa ramete rs of the las t least-squares calculation are also listed in Table 2.

Table 3. The assumed coordinates of the hydrogen atoms at C(7) and C(10)

x y z tt(7) 0.287 0-180 0.232 H(8) 0-191 0-192 0.332 H(9) 0.225 0.392 0.286 I-I(10) 0.874 0.213 0.326 H(ll) 0.959 0.386 0.421 H(12) 0.989 0.158 0.452

I t is clear t h a t the electron dis t r ibut ion around C(7) and C(10) cannot be represented by single positions of three hydrogen atoms. An approximat ion for these electron densi ty distr ibut ions was therefore made by pu t t ing six hydrogen atoms, wi th half weight, a round each of the two carbon a toms in the manner indicated in Fig. 2(a) and (b). These locations were chosen with a C - H dis tance of 1-02 A, which is the average value found for other C--H distances in this s t ruc ture determinat ion. The calculated coordinates of the six hydrogen a toms are l isted in Table 3.

One addit ional s t ructure-fac tor and least-squares calculation was done with the coordinates l isted in Tables 1, 2, and 3 as input parameters . The six hydrogens with half weight were assigned a temper- a ture factor B =4 .0 j~2. This calculation did not show any improvement in the agreement between the observed and calculated s t ruc ture factors compared with the previous one; no significant corrections to the positions of the heavy a toms were found. The s t ruc ture factors obtained f rom this calculation together with their observed values are l isted in Table 4.

Attempts to refine the y coordinates

To this point the structure has been refined on the basis of the space-group s y m m e t r y P21/m, so t h a t all the carbon, nitrogen, and oxygen a toms were lying in the mirror plane. Rela t ive ly small deviat ions of the a toms from the mirror plane, however, m a y occur if the planes of the two molecules do not coincide with (010) or if the bases are not p lana r molecules. Such depar tures of the a toms from the mirror planes would be incorporated in the anisotropic t empera tu re factors fli. The principal axes of the thermal ellipsoids (Table 8) normal to (010) are in all cases except for C(6') and N(9') the largest ones.

T a b l e 4. List of structure factors for A T

T h e t h r e e c o l u m n s of e a c h g r o u p c o n t a i n f r o m le f t t o r i g h t t h e v a l u e s of ]c, lOFo a n d lOFc. S t r u c t u r e f a c t o r s l i s t ed w i t h a n a s t e r i s k w e r e n o t i n c l u d e d in t h e r e f i n e m e n t b e c a u s e t h e y w e r e n o t o b s e r v e d or w e r e b e l i e v e d to su f f e r f r o m e x t i n c t i o n

ool

O* - 3o~0 I* - -98

9' 58 -59 s 1.9 63 6 25 -26 7 68 69 8 15 9 66 -63 io ~1.~ ~ 9'2 _1~

15 1.4 -12

IOZ

O* - -622 1. 59 -6s 2 1~7-161 3 loS l l s 9' i~ -1.52 5 11 i0 6 276-290 7 55 --52 8 82 -92

11. -]2 i0 91 -95 11 70 -7~ ].2 <10 1.3 1.8 50 11. 52 -42 15 27 28'

1 h3 -51 2 205 -219 3 68 69

18~ -18~ 5 120 -116

35 -45 1.o3 111

81o6111

11 ]21. ;i.26 12 ~0 20 13 • i0 1~ < 9 " 57

20 l

"-323 o 2~ 1~ 2 i07 3 52 56

-200 29'7

59 135-137

98 lo~ 9 163 -166

13 29 -5 14- 19

73 -7"5 2 21 21.

-210 5 31 -~5 6 28 -2o 7 1.77-185 8 < 7 -17 9 l l l l

i0 ]22 -133 ii 112 -123 12 39 -33 ]o ~ -1.3 i~ i0 15 -11

3oi

o~ 1.7 -129

2 ].oh 92 ~ 112 -12o

29Z 280 1.9 -13

58 lO1 -101.

190 37 32 29 26

13 9 -9

1 29 26 121 -135

3 72 -73 2 ~13 lO 5 136 1.1.8 6 lO5 m-3 87 ~_~8 -9o 9 23 1.~

i0 11.

12 1.3 21. 2~ 11. <9 15 < 8 -5

1.ol

0 2 ~' 27 1 1.'.8 I39 2 ~ 61 3 ]28 4 ~ -21 5 1 -11.3

76 ~ - ~ 8 <12 -6 9 5o 32

1~1° ,297-~ 12 27 -25

~oI

1 ]2 -16 2 10 -17 3 259-271 4 21.0 21.9 s 3o6 -3~ 6 <io -3 7 75 71. 8 i0 -7 9 22 -31 .o ~3 -47 .1 ~7 -4o 2 <9 -7

L2 LS 9 -6

50l

0 h6 -h} l 71 73 2 25 23 3 30 28 9' ].18 1.13 5 156 lhO 6 86 88 7<11 8 <1o i 9<9 9 .o ]3. -16 1 27 -22

- 50~

1. 17 21 2 < i0 -2 3 5D 50

31. 38 5 <9 -1 6 ~ -2o 7 lO9 if3 8 < 1 o 2 9 2 ~ 1.89

LO 1.3 -51 i-1 83 81. 12 78 -77 ~) 1.o -39 t1* 36 27 ].,5 1.2 -32

60l

0 ] .?7-]-96 i 1.1. 22 2 135-137 3 lh~ -150 9' 1.6 16

95 -96 62 6O 52 -61 9 i0

9 • 8 -2 i0 "52 -40

I 72 73 2 23' 1.6 3 51 -47 1. 41. 3~ 5 108 105 6 i~1. 128 1 9'7 h2 E 9 ° -E S -36

1c 58 55 35 -33

i] ~.o 3~ i~ ~ -36

3~

7oi

0 52 -50 1 71 -77 2 ~ -16

43 ~ -36 5 <i0 5 6 <9 lO 7 ~'2 36 8 14 1.3

9o~ i 69 -69 2 12 -16 3 16 22 1. 31. 31 5 109 -1O6 6 5~ -51. 7 16o-167 6 ]-I -3 9 61. -63 IC i0 .-8

12 i0 7 i~ < 9 2 111 < 7 2 15 14 -11.

801

82 88 23 22

2 <9 -8 26 23

< 7 -8 5 12 -]2 E ,4 -1.3

1.3 1.7 2 i0 -4 5 • i0 11. 1. < lO -2 = 95 1.07

1.1 -4E 99' ioi ]2" -c <9 -3

iC 21 .1~ ~1 9 1,~ 29 -25

9Oz o ~6 hl ' 1 35 -3o 2 27 25 3 32 -28

_ 9OZ

1 22 -26 2 32 27 3 I0 -i0

26 -31. 17 -19 34 31.

7 ( 9 -.5 2~, 21.

9 1.8 -22 LO 13. 6 L1 26 25 U~ 13 -1.7

0,0,1

o io -8

6,0,1

i <8 -k 2 <8 8 43 s8 52

9 -]2 21 2~ 9 5

7 21. -23 '¢8 9

9 1-1- 3

01.l

l * 21.2 273 2 38~ 388

3 <9 -1. 306 -322

~ ]2 -4 16o -151

7 6,5 6o 8 <1.1. 1. 9 96 98 io <11 9' 11 9'1. 1.~ ]2 1.8 ~9 13 3o 26 19' l l -12 15 1.6 Ii

111

0 21.1 24h

2 32 h2 3 17 -27

16 15 5 252 229 6 91. 78

21 -19 18 -21

9 3,52 165 10 97 -98 11 11 9 ll~ 9'9 -~8

].6 1.6 15 18 -20

_ l l Z

i 23 -26 2 --6O9 3 ~5 -~3 9' h51 289 5 399 - 4 ~ 6 25O 21.7 7 131-]23 8 26 53 9 ~9 -47 I0 i0 7 ll 66 66 12 38 31. z3 1.8 Z2 9 -3 ].5 27 -20

21l

o 51o 3z9 i 288-28~ 2 65 70 3 231-237

]26-1.69 5 192-195 6 24 -21 7 21 -16 8 132 131 9 116 -]2~'

lO ]27 1~3 ii 1.3 -i0 12 12 13 13 35 32 14 8 -7

1 iii-122 2 ~2-367 3 1.95-197 5~ 9'6 ,-~8

• 97 -97 6 lO8-1o5 7 16. -o 8 78 -79 9 1.1. 16

lO 35 ~ 7 38 -40

12 39 -36 13 95 -Ill 1.1. 1.9 53

311

o 37~ 371 1. 99 95 2 1.63-156 3 202 206 2 2o~-2O9 5 42 46 6 75 -79 7 13 8 8 91. -97 9 8~, -91

12 -i0 13 ]2 1.o

I 13 6 2 221 21.9

9' 30 39 5 171 1.56 6 2h7 237 7 21.7 2o6 8 <9 -]2

LO 9 39 1.0 95 -105

u. ~ -36 m 18 -21 L~ ~ 1.o L4 39 -h9 L.5 ].8 -9

o 197 -181 i ]-1.3-1~2 2 77 68 3 lO 11

6 117-Lt6 7 1/2-1/i 8 15 11. 9 1.8 22

1o 85 1.2 u. < 8 -4

i ]26 ]..19' 2 ?-1.7 226 3 22 -21 9' 21.2 257 5 1 1 5 - l O ~ 6 17 -]2 7 ~9 -48

72 -66 9 ].32 13o

zo 75 -72 11 21 23 ]2 ]2 -1.5 zD 29 31 1~ <8 2 ].,5 57 2~ ].6 21. 22

51Z

o 67 i 87 2 202 198 3 17 -11 4 31. 35 5 ~.6 13o

<8 .{~ 1.o 19

glz

1 s4 -~ 2 136-]28 3 73 75 9' ]2 -ll 5 88 8.5 6 9'7 -56 7 75 -71 8 125 13.8 9 150 -.159'

io 181 189 L~. <9 2 ]2 3~ 33 13 28 -27 11. < 8 3 15 2O -12

61l

0 100 98 1 27 ~30 2 9 15

27 -28 ~ - 8 -68

~8 6 8

glZ

i 76 77 2 9'6 -46 3 72 75

139 -1.~ 5 ~2 -40 6 ~/~ .-5o 7 9'5 38 8 22 -17 9 9'2 -46

10 77 -73 11 20 -21 l ~ 21. -26

18 17 11. 2O -1.8 15 26 23

7].l o 353 -162 1 58 62 2 92 -98 3 33 35

9'3 ~7 5 z9 -5o 7 <8 6

<8 -3 8 18 -13

i 2o~ -~11 2 1.06 ii0

26 21 122 134 1.4 9'3

7 37 1.3

9 1.9 -i0 i0 8 5 : 1.% 14 -]2 ]2 ]2 -8 13 3o 22

81.Z

o 56 -36 i <8 I 2 <8 -2 3 <8 -2 1. 37 23

31 20 1.8 35

81.1

1 lo9' L u <9 1

3 29 2 8 9' 37 -38

26 -25 22 -23

7 26 -22 8 9 0

lO 11 -19' ]2 12 -7 1.3 8 3

91~

0 8 -7 i <8 -i 2 <7 5 3 53 -42

§1.z

1 22 21. 21. -20

3 "2O 25 9' 10 ~ ~ 18 -

35 -31. ~ 2o

g ~ 09 26 27

12 20 25

1 ~7 -1.7 2 g~ -_~

<8 -5

- - 1 ~ 2 O' 9'9 6o 2' ~i0 h53 3 25 -~8

3o zo6]" -665

7 51 -45 8 ~ -3 9 51

lO 9 -3

27 -20 1..5 1.5 ]2

12l

2 138 139 3 68 -72

88 92 1o -7

2]2 211. 7 1.2 ~5 8 75 75 90 17 13

78 81 ii 62 61.

<8 -6

36 38 159' 151. 75 -8~

]-17 120 86 85 31 32 75 -76 77 -8~ 68 .-6~

iC 160 -155 i/ 115 -i08

Zg < 7 -3

221

203 222 ~5 -49

-38 9'2 -40

1.72 "138 209 -191. 57 -50

ZO9 1.07 89 -9O

11.1 3.39 lC 53 - ~ 1/ 9'5 9'8

37 -3o 1] < 8 5 1.4 18 -17

36 24

2 Io~

3 "~,3 L~2 120

15 18 7 1.56 1.50 8 1.o 9

"I 103 105 31 28

L3 i0 ]2 ~9' .8 -9 U5 19' 9

321

o 6]. -35 1 112 103 2 69 -58 3 88 93 1. ~o-2o9

-25 7 9'7 -~8 8 89 85 ~ ~-28

-20 .i 3~ -31

-3

52l 1 79 -53

113 105 3 ~5 39 9' 9 ..o

11o -lo9 85 -81.

7 57 ~6 8 77 66 9 35 -31. LO ].9 -]-9

73 -7o

L~ -17 ~9' 7 9'

9'21

o 17 -19 1 1.19 -1 /1

67 -57 3 97 -92 1. 25 26

~ 66-~I < 8

LgO 31. 28 "2~k

,1 < 9 2 2 11 1.3 3 200 207

:1.~:~ -172 5 261. 259 6 < 8 5 7 6o -s5 8 <8 2

37

<7 7 ]-3 31. 33

52 l

0 32 27 58 -59 25 -21.

3 20 -21 4 98 -95 5 ]23-L~7 6 78 -73 7 <9 -3 8 <8 -1

g2~ 20 -21 <9 55 -43 16 -1.9 ~ ~ _1

1.3 7 83 -88 8 15 -9 9 165-I,57

io 1.0 42 11 7"/. -.65 12 63 65 ]J ~ 33 lh 3o -22 ]..5 33 28

621

lC 176 159 9'i -38

].21 117 129 125

~ ]2 -13 85 ~ 9,7 -4E

S <7 3

621

61.. -6] ; 30 -~ : 37 [ 3o -2.= , 90 - ~

119 -]2.= 39 -3~

i

1: 29 ~.. 1., 38 -3 ¢. 1: 29 -~. li 51. L, 30 -3:

721 0 ~5 9'2 1. 63 66 2 i0 i0 43 55 -5~

28 29 5 <9 -5

9 -7 35 -31

1 61. 58 2 18 18

19 -22 43 26 -2~

50 7 139 1~1. 8 9 3 9 52 55 LO 9' 8 ~i 31 27 12 8 -6

o 71- -73 1 22 -1.9 2 < 8 43 27 -19

<7 6 5 ii i0

1.5 -11. 2 9 3 3 < 8 4 - I 1. <8 1. 5 87 -92 6 ~ 39 7 86 .-86 8 <8 -0 9 <8 2

LO 17 .~ L1 <7 [2 24 21

92~ o 9'1 - N

32 26 2 22 -22

1 21 21 2 29 -22 3 ]2 lO

25 26 ~ 1.5 16

27 -27 7 <8 5 8 24 -20 9 18 ].8 kO i0 -6

__ I0, 2, l

i <8 2 < 8 -~ 3 55 -46

8 9

9 -~

1 199'-183 2 193-192 3 22 -25

Z6Z 172 5 < 13 -5 6 1o7 io~ 7 47 -42 8 zo -6 9 62 -63

10 17 -19' u. 29 -27 ]2 3~ -37 1~ 25 -19 11. 9 7

131

o 133 -132 1 51 61 2 i0 -4

32 -27 37 -5O

1.39-11.3 50 -~8

~ 11 9 23 22

9 Lt3-1~3 lC 69 7~ ]J 9 L~ 37 38 i ; ~o 35 lzl 11 -1.l

_

] 19' -15 ,~ 262 281

17 1S 232 -227 228 2 ~ 19'3 -1.3~ 92 8] 26 -~ 37 3~

1.( 9 -~

i.. 27 -2'] 1.3' 17 -1.~

231

1.~6 -1.3; 1.61 .15~ 31. -2~ ]23 Ii" 102 9" 122 11.= 1.4 ~

91

91 -96 8

12 14 -]2 L 3 26 -23

~3Z z 6o s8 2 176 183 3 88 9O 1. 5~ 58 5 ~7 1.8 6 ,"8 75

-1.

9 3.8 -15 i0 ~ -26 11 31 12 3o 28 13 89 8z 19' 36 -36

0 169-16s 1 71 -6~ 2 92 91.

135-]25 138 1.33

39 k l 7 1o -1 8 66 67 9 6o 6%

1 1.5 6 2 ] 21 . - i 0~ 3 67 75 ~ 9 -8 6 150-19'~ "7 ] .22-135 8 lO -2 9 2o -2o

1.2 13 ].5 ~3 31 -3O L1* 55 D~ U5 13

o ~3 119 98 88

2 9'3 -32 9 -5 65 -55

7 82 82 8 11 -9 9 1.1. -16 m :,8 -32

1. 81 -72 2 122-122 3 < 8 8 4 i~2 -I~O

63 55

8 51 50 9 9~ -91

51. 54 17 -19

1~. 8 -3 15 27 -18

53z o~o , .~ 2 120 -113 43 8 2

22 -22 91. - ~ 27 -26

7 6O -6) 8 8 -7 9 < 8 10 14 13

g3Z

9'°_~ 211.7 i 43 8 5 67 -63 6 39 36 ~ ~9

-81 9 lO3 zo~

lO 1.32 -13~ 11 < 8 -1 12 27 -26 13 < ~ 1.8 19' -3

631

~ 89 -7~ 19 ie

2 9 -'9 43 18 IE

19'

< 7 -3

1 62 -5E 39

94 9-' s ~ 3~

3o 3:

53 55 17 18

].2 ]-9 .17 ~3 13 -]2 i~ :]-3 13

73 l

1.1 -43 2 68 7O 43 21 -23

29 -31. 21. 21 <8 -a

7 6 S

159 1.50 77 -77

43 2o 17 26 -22 92 -98 33 -29

7 25 21. 8 1.3 ]2

.1~ "~ -5

1~ 17 -14

831

o 27 21. < 8 _1

2 <8 o 3 <8 -2 1. 21. -L5

1 78 -79 <8 -2

3 22 -21. 9' 30" 26 5 21 20 6 ~o 20

:]-3 ]2 ~ 8 2

9 <8 2 zo <8 -~ i i lO

1 22 -16 11 11 21 -1.7

6 22 7 17 -14

24 -21. 9 7 -8

O~ 9~2 72~ 1 15 - ] 2

9'5

8 11 2 9 3O -28 io < 8 -1 1% 18 19 [2 25 -21

].9'z

o 220-220 1 ]-3 -1-I 2 2~5 -72

26 43 27 -~5

8 -11 L05 -i03

7 26 -29 ~_~8 -10

i0 115 J~9 ].3. 37 -39 ]2 "=7 2 13 25 26

1 15 -].5 2 64 -73

~i -41

-26 7 9'2 4O

37 9'2 9 9'9 1.6 lO 89 92

69 7o ]2 28 2~ 1.3 8 6 1.4 7 -1

24~

]2 16 X ~ - ~ 26

9'5 97

61 -6O

lO 3o 35

13 ..k 1.8 - 8 24 24 s9 -57

s 38 - 9 6 8 - lO 7 81-83 8 <? 0 9 1.7 -]2 io 61 -61 11 67 -7o ]2 23 -18 13 9 -1o 1.1.. 8 7

0 23 z~ 1 55 --~ 2 18 16

101 i01

%~ 15-7 7 27 26

5~ -5~

~o ~ ~9 1. 11 9

NI

21 32 56 -51 43 <8 -4

< 8 -s 55 62 9'5 9'3

7 25 -2O 9']. -~6

9 14 13 10 18 18

]2 9'2 ~1. 13 12 /_%

0 15 12 1 56 57 2 43 38 3 ~6 ~. 9' 25 -25

67 -71 1.i -39

7 37 36

lO :].8 -1.3

~z

1 ~ -8 2 -7 3 1.o9 -io7

75 75 ~36-1.35

7 197 -7 28 ~ 8 -4

12 -13 110 ~:~ -22

27 -25 ].2 7 -6 13 21 -18

~z

o 9 -3 1 ~3 a6 2 18 20 43 1.9' ~

55 58 7~ 79' ~0 9'3

1 1.8 17 2 9 5

< 8 6 <8 3 11 -7 h3 9'6 16 1~

19o 91. 9~ 29 -27

11 9'3 39 ]2 52 -42 13 30 -2O

6kZ

9O -93 27 25

2 75 -76 3 7o -76

~2 -56 6 30 26 7 32 -31

1. ~0 9'0 2 9 1 3 20 -:].8 1. ]2 i0 5 54 56 6 68 72 7 26 29'

1.o ~ 33 11 17 -17 ]2 27 28 ]-3 15 18

71.l o ~ -27

-41 2 <9 -2

35 33 ' 7 -:L5

1 ~ -37 2 -19'

14 5 51 -57

8 7 -5 1 7 61 61

1.: ~ - ~ 1 ~ .' u . 17 -15 I 10 38 -~6 ~ g - 1 ,

~z -7 o 1.6 9's 1 13 17 18

1 16 13 1 ~.5l 2 <8 .L, 3 13 1-I I o 54 -57

1 ~7 -48 89'7. 2 20 16

lo, ~' 3~ ~,

9' - o , 54 ~ 7 53 57 1 8 6 3

<7 o a ~ Z

9 17 2 , 2 1 1.7 9'1 1o -9 I 9'7 51

0 1.1. 2o I 5 28 -25 6 <8 -4

~z 7 <8 -8 8 33 -32

1 9 - ]2 I 9 49 50 11. :]2 I lO 32 -31 I0 -9 I i I ik 13 lk -15 I 12 io -8 1o -]..I. I 1.3 6 8 2o 17 1

7 9 -5 ' 55Z 8 1.6 ]2 I

lO 25 25 33

2 45 9'8 05Z 3 < 8 9'

1. 10 9 62 67 1 5 58 5~ 89 871 6 15 15 11. 1~ 1 7 ~5 9'O 63 -681 8. <6 1.

9'9 -~71 gSZ 28 29 I

8 9 -, l 20 -19 32 50 1 2 37 -)7

i0 17 15 1 .5 25 28 15 121 4 7 -3

]2. 20 ~-~ I 5 32 3z 6 L5 -16

15 l 7 28 -27 8 1.2 23

1+5 -29 1 l o 76 75

2 ~8 - O a l l 7 2 43 19' 12 I 12 15 16

27 2~, I 56 6O I 6.5Z 1.9 15 I <9 o = 0 9'2 9'2 ]-9 -18 1 ~ 7 -8

0 58 6o I 2 9 "7 1. 9'6 -J~ 1 3 8 -10 11 7 ~, 9' i° -7 ].2 2 / -2~ I 5 36 -31

6 <6 2 i=,z ~ z

21 9 ~ , z 35 3' ~17 -116 1 2 2O -2:

3 12 -10 I 5 15 1. 81. 82 I 9'3 -4~

88 -91 I 5 21. -2.' 68 6o I 6 14 -1(

7 ~3 -41.1 7 11 <8 .~ l 8 8 --']

9 24 -23 I q 12 -1.= io < 8 1o • 29 -3.= l l 27 2) I 1.l 9 -~-" ]2 19 16 1

o 66 ..611 ~51 i 22 21

0 2 1.3 -3~ i 6~ -59 3 i0 10 2 i0 ~ I ~, 13 i~ 3 61 -601 4 ~o ~o i ~SZ

..~ , 45 1.0

9'9 ~ I 22 18 59 52 ~ ,58 55

2O 1.6

~ -21, i 2 --62 I 3 33 -31 I 9' 122 -22 l 06l 5 15 -151 0 291 -268

~8 -451 1 (i0 _4 10 -z ' 2 76 -8~ 16 -i~I 3 <9 2 13 io I 9' 9 6

io io 11 1 5 20 -20 ~-18, 11 ~ 0

-17 I 17 -17 • , ~9 -~,~. ~ _2

l l 35l

C 5~ 56 I 16Z 1 ~0 ~0 I 0 92 9 ) 2 35 ..-58 1 1 < 8

57 5 8 1 2

<7 ~ ~318 39 -',~ I 1.~ ,7 -,,, 87 2~ 1.1.

25 2.5 -22 1 9 5

5~z ~ 23 23

I ~ 14 -I0 I .[.6l h5 42 I

43 ii -12 9 9

5 < 8 6 6 22 2O 7 20" -21 8 1.1. -16 ~ ~ .~8

~ 3 11 1.5 -36

261

o .28 27 1 7 -63 2 8 3 16 -'1.7 1. 26 22 5 38 -42 6 1.7 -18 7 32 27 8 37 -32 9 1.1 36

'~6z

1 9 5 2 <8 -1. 3 < 8 - 9' 20 211 5 21 2~ 6 < 8 7 31 36 8 7 -3 9 1.7 13 1.o 31 3o 11 9'3 39

361

0 7 "~ 1 1.3 11 2 <7 -0 3 12 15 4 ~0 -38 5 <7 ~ 6 <7 7 ].3 - ~ R ~1 o-,

I 8 -]2 2 21 18 3 <8 -6 9' <8 2 5 ~ - ~ 6 18 -21 7 9 ~ 8 22 9 < 7 -3

10 13 -12 11 28 -22

o 7 -7 1 9_I -22 2 22 -20 3 16 -18 1. 13 16 5 39 33 6 ].8 18 7 17 -16

E61

l 1~ IS

2 <7 k.9~ 3 1'7 9' 28 -28 5 52 57 6 <7 6 7 20 -ii 8 6 7 9 <7 6

i0 12 i0 L1 :]-3 :]-3

561

o <7 1 <7 2 13 -13 3 ]2 -]2 9' 27 -28 s 38 -38 6 11 -2o

1 9 -lO -io 2 9 -~

3 ].7 4 < 7 -~ s <7 6 < 6 7 16 -19 8 l l -lO 9 ~ -m~

i0 14 12

661

0 ~3 ~3 1 20 -13 2 ~3 9'0 3 35 36

g61

21 22 _2~ o 3 9 8

< 8 -3 5 25 -28 6 28 -33 7 12 -12 8. <7 -3 9 <6 15

761

o 19' i~ 1 8 20

1 19 19 2 7 7 3 13 -lO 1. 6 -4 5 28 3O 6 14 15 7 26 47

o '~ 1 l? -23

17 7 18 -16 8 8 -6

17~

~ 2s j j - ~

2 <7 -i , ,~ .~

7 <7 -2 8 9 i.%

in

21~ 4 3 8 7 9' 21 -26

-21. 7 2O 8 <7 2 9 <6 13

27 l o ~ -12 1 2O 2 <7 i

11.

271

~ 9 8 7 6

5 <7 2 6 29. 2O

3TZ 0 16 -16 i 25 -22 2 16 13 3 26 -2~ 9' 26 24 5 10 -8

~ ~ - ~ -28

7 27 -2 t 8 7 -7

271 0 22 2% i 24 21.

9 -7 3 7 -1 9' 19' -13

1 25 -21 17 -z8

15 io <7 2

87 "%' 1' 57Z

0 i0 -13 1 ~ -]5 2 -16

771 lk 10

-13 43 <5 .-O S ~ . -13

o81

0 113 9~ l 7 5

18z

0 38 -38 l 11 -3

3 7

28l

3

1 < 6

-2

58*

912 COMPLEX B E T W E E N 1 - M E T H Y L T H Y M I N E AND 9 - M E T H Y L A D E N I N E

Although this feature is common for a layer structure, some of these axes, especially those of the annular substituents, are larger than could be expected from temperature movement alone, indicating that devia- tions from the mirror plane might exist. Furthermore, the observed electron distributions due to the hydrogen atoms of C(7) and C(10') (Fig. 2(a) and (b)) cannot be described by single positions conforming to the symmetry restrictions of the mirror plane. The question now arises whether or not those deviations from the mirror plane are ordered in the sense that they conform to the symmetry operations of the twofold screw axes. In the first case, a better descrip- tion of the structure can be given by assigning it the space group P21, but in the second case no real advantage can be obtained by the omission of the mirror plane; the deviations in this state occur in a disordered way throughout the crystal. In order to find an answer to this question, the refinement was continued with the calculations based on the space group P21.

In order to remove the mirror plane from the space group P21/m, some atoms had to be given y coordinates different from ¼. In the structure of adenine hydrochloride (Broomhead, 1948; Cochran, 1951) the z coordinates were calculated on the basis of a planar molecule, so that no deviations from planarity were known for the 9-methyladenine mole- cule. For the structure of 1-methylthymine (Hoog- steen, 1963) the following departures from the best plane of the molecule were found: N(1)=0.014 A, C(2)=0.005 A, N(3)=0.005 A, C(4)=-0 .009 ~, C(5)--0.002 J~, C(6)= 0.009 A, C(7)= -0 .035 A, 0(8) = - 0 . 0 1 7 J~, 0 ( 9 ) = - 0 . 0 3 5 /~, and C(10)= 0.028 A. The atoms C(7), 0(8), 0(9), and C(10) were assigned y coordinates that gave the same distances for these atoms to (010) in the AT structure. In the following structure-factor and least-squares calcula- tions only the positional and temperature parameters of the heavy atoms were refined. The weighting scheme used in the final stages of the centrosymmetric refinement was applied. The calculations were termi- nated after six refinement cycles, since no shift was larger than 0.03 times the standard deviation.

A comparison between the coordinates obtained in this way and those of Table 1 showed that none of the final parameters of the centrosymmetric refine- ment was significantly different from those acquired by the aeentric refinement. The largest shift in the y coordinate was for C(7): 0.061 /~, with a calculated standard deviation of 0.01 /~. Although the dis- agreement index R decreased from 0-081 to 0.073, the sum of the squares of the weighted residuals, Zw[/IFg] 9', remained constant. Since the acentric structure contains 63 parameters more than the centrosymmetric one the standard deviation for an observation of unit weight actually increased, indicat- ing that the acentric refinement did not converge to a set of coordinates significantly better than those

derived from the refinement with space group P21/m. The following discussion is therefore based on the parameters of the centrosymmetric structure listed in Tables 1, 2, and 3.

Discuss ion

In Fig. 3 a view of the structure down the b axis is presented. The molecules in heavy lines belong to the same layer. Part of an adjacent layer, one-half of the b axis down and related to the upper layer by the twofold screw axes, is also shown in the top half of the figure. The structure is a layer structure with

0 .X / ', / /

1.,.~. 4"1

2' . . . . 1

-u; ~ :- i

V °,,a ..... ~¢0 /

Fig. 3. A view of the s t ruc ture viewed down the b axis. The small circles represent carbon atoms, the large single circles oxygen a toms and the double circles n i t rogen atoms.

layers of base pairs parallel to (010) stacked over each other. The amino group of 9-methyladenine is hydrogen-bonded to 0(8) of the 1-methylthymine molecule that lies one unit cell up along the negative direction of the a axis. Each layer is thus formed by columns of base pairs running parallel to a. The cohesion of the structure within each layer is provided only by the van der Waals interaction between these columns.

The three methyl groups at C(7), C(10), and C(ll ' ) are pointing towards each other, with the following distances :

C(7)-C(11')=4.19 J~, C(7)-C(10)--3.74/~, C(10)-C(11')-- 3-52 ~ .

There are no short intermolecular distances between molecules in the same layer. The packing of the base pairs within a layer is shown in Fig. 4.

Since the b axis is 6.552/~ in length, no interatomic distance between molecules of adjacent layers is shorter than 3.276 /l. The shortest distance across two layers occurs between C(4') of 9-methyladenine and C(2') of the molecule related to it by a twofold screw axis; it is 3.285 A.

Interatomic bond distances and angles The interatomic bond distances and angles of the

base pair are shown in Fig. 5. In Table 5 are listed

K A R S T H O O G S T E E N

z - C

7 - - . • . , " . . . ./. ..

x ,' :~ . . . . . . . . . ~ . . . . . . . .,,~,., . . . .. ~ - . " . - . . , . , . , i t . . . . . . ~ . .

• , . , .. ~- ..~ , , , . #', . ,~ .. t .:, ~ . / . • ~ I " " " ~ ~ " " " ' "

• "* :... -:..~ .. , :...,-. .',, . .,,... : , • ,. ..-. • • • . , : ~ . , - ..-,,~ ,,r , '-'

• ' ' ; ,..:.i.,k~".:~;. ~ : . , , i : . S . ""< ..: ; ," :'~ ~::~: .,:,;.'.;.~' . .!:,:<~:~7, ,<"- ' - , . I~:.': . , , ~ . j •

Fig. 4. Pack ing of the base pairs in a layer p a r a l e l to (010).

913

9 1 . 3 " - 0 . 8 7

~422/2

/ \ ,.sssk?~,~o - ,,5.7-J.~ -

,6"~/Z - ~ \ V f - - , 8 6o - ~ 9 ~ 0 . 9 7 ~ ,

_

~//~,,~.,- , ~ . L _ ~

1"40612~(F>/I\~~/~~"/8'0°

132. I°,,~.,s .~ 116"9"

,~ ~o5.5 o i 0 4 . 4 o ",.,

124 ,>

0 " 9 2 t )

130 ,9 •

127.1 °

125 .6 °

4 8 . 4 °

7 1 2 6 o . , / "

1 - 3 0 4

IO3

H i 4

Fig. 5. Molecular dimensions of the base pair.

the dimensions of the 1-methylthymine molecule as they occur in the crystals of the AT complex, together with the calculated values of their estimated standard deviations. These values together with those of Table 6 have not been corrected for tho effect of thermal motion. The dimensions of 1-methylthymine as deter- mined from the structure of this compound (ttoog- steen, 1963) are also listed. A comparison shows tha t

the bond lengths of the two structure determinations are in good agreement. The largest difference, 0-014 for C(7)-N(1), is only of possible significance. No differences are found in the internal bond angles of tho pyrimidine ring. Larger discrepancies, however, occur among the external bond angles which seem to change more with different structural environment than the internal bond angles or bond lengths.

914 C O M P L E X B E T W E E N 1 - M E T H Y L T H Y M I N E A N D 9 - M E T H Y L A D E N I N E

Table 5. Bond distances and angles of 1-methylthymine AT: this structure determination

NMT: 1-methylthymine (Hoogsteen, 1963)

AT NMT /I zJ/a(gl) ( o = ( o =

0.005 A) 0.004 A )

N(1)-C(2) 1.376 h 1.379 A 0.003 A 0.5 C(2)-N(3) 1.378 1.379 0.001 0.2 N(2)-C(4) 1-377 1.375 0.002 0.3 C(4)-C(5) 1.422 1-432 0.010 1.6 C(5)-C(6) 1.333 1-346 0.013 2.0 C(6)-N(1) 1.382 1.383 0.001 0.2 C(7)-N(1) 1.456 1.470 0.014 2.2 O(8)-C(2) 1.207 1.214 0.007 1.1 O(9)-C(4) 1.238 1.237 0.001 0.2 C(10)-C(5) 1.510 1.497 0.013 2.0

( 0 = 0 . 2 ° ) ( 0 = 0 . 2 ° )

N(1)-C(2)-N(3) 115.0 ° 115.4 ° 0.4 ° 1.1 C(2)-N(3)-C(4) 126-3 126.3 0.0 0.0 N(3)-C(4)-C(5) 115-7 116-1 0"4 1.1 C(4)-C(5)-C(6) 119.0 118.3 0.7 1.9 C(5)-C(6)-N(1) 123.2 123.3 0-1 0.3 C(6)-N(1)-C(2) 120.4 120.6 0.2 0.6 C(7)-N(1)-C(6) 120.1 121-2 1.1 3.1 C(7)-N(1)-C(2) 119.4 118.2 1-2 3.3 N(1)-C(2)-O(8) 124.3 123.3 1.0 2.8 N(3)-C(2)-O(8) 120.7 121.3 0.6 1.7 N(3)-C(4)-O(9) 119.9 120.0 0.1 0.3 C(5)-C(4)-O(9) 124.4 123-9 0.5 1.4 C(10)-C(5)-C(4) 117.3 119.3 2.0 5.6 C(10)-C(5)-C(6) 123.6 122.4 1.2 3.3

Table 6. Bond distances and angles of 9-methyladenine AT: this structure determination

A+HCI': adenine hydroehloride (Cochran, 1951)

AT (o = 0.005 A) A+HCI"

N(I ')-C(2') 1-361 • 1.37/~ C(2')-N(3') 1.304 1.30 N(3')-C(4') 1.347 1.36 C(4')-C(5") 1.373 1.37 C(5')-C(6') 1.406 1.40 C(6")-N(I') 1.355 1.38 C(5')-N(7') 1.381 1.37 N(7')-C(8') 1.323 1-35 C(8")-N(9') 1.363 1.33 N(9")-C(4") 1-389 1.36 C(6")-N(10') 1-335 1.30 IN(9")-C(11") 1.453 - -

(c;= 0.3 ° ) N(I ' )-C(2')-N(3') 130.9 ° 124.5 C(Z')-N(3')-G(~') 110.0 112.5 N(3')-C(4')-C(5') 127-4 128.0 C(4")-C(5~)-c(6 ") 116.9 118.0 C(5")-C(6')-N(1') 118.0 114.0 C(6")-N (1')-C(2") 116.8 123.0 C(5')-N(7')-C(8') 104.4 102.0 N(7")-C(8')-N(9') 112.9 115-0 C(8')-N(9')-C(4') 106.2 105.0 • N(9")-C(4")-C(5") 105.5 107.0 C(4')-C(5')-N(7') 111.0 111.0 N(10")-C(6')-C(5") 123.0 126.5 N(10')-C(6')-N(I ') 119.0 119-5 C(8')-N(9')-C(11') 128.2 - - C(11')-N(9')-C(4") 125.6

A discussion of the dimensions of 1-methylthymine in terms of valence bond structures has already been given for the structure of 1-methylthymine and this discussion seems to apply to the molecule as it occurs in the structure of the base pair as well. The dimen- sions of the 9-methyladenine molecule are listed in Table 6, together with those of adenine hydrochloride as determined in the structure of this compound (Broomhead, 1948; Cochran, 1951). The standard deviations in the bond lengths for the latter structure were estimated not to exceed 0-01 /l. A comparison of the bond lengths of this molecule shows that the addition of a proton to N(1) apparently does not alter the bond distances appreciably. Larger differ- ~nces are found in the bond angles. The angle C(6)-N(1)-C(2) is 116-8 ° in 9-methyladenine, whereas it is 123-0 ° in the adenine ion. The attachment of a hydrogen to the N(1) atom apparently increases the ring bond angle by about 6 ° . This difference is in good agreement with the one given by Pauling & Corey (1956).

The most important feature of this structure is the manner in which the two molecules are hydrogen- bonded together. 0(9) of 1-methylthymine is hydrogen-

Table 7. Distances and angles involving the hydrogen atoms for which the coordinates were determined from

the final difference-Fourier synthesis C(6)-H(1) 0-99 A_ N(3)-H(2) 0.97 N(10')-H(3) 1.08 N(10')-H(4) 0.87 C(2')-H(5) 1-01 C(8')-H(6) 0.92 C(1 l ')-H(13) 1-00 C(1 l')-H(14) 1.03

C(4)-N(3)-It(2) 109.3 ° C(2)-N(3)-H(2) 124.3 C(5)-C( 6)-H( 1 ) 116.4 I~(1)-C(6)-H(1) 120-5 C(6')-N(10')-H(3) 133.0 C(6')-N(10")-H(4) 122.2 N(I ')-C(2')-H(5) 102.9 N(3')-C(2")-H(5) 126.2 N(7')-C(8')-H(6) 123.9 N(9")-C(8")-ti(6) 123-2 N(9')-C(11")-H(13) 108.5 N(9')-C(1 l ')-H(14) 104-3

bonded to the NI-[2 group of O-methyladenlne with N H - - - 0 = 2 - 8 4 7 _A_. N(3) of 1-methylthymine is hydrogen-bonded to N(7) of 9-methyladenine at a distance of 2.923 A. In a review given by Fuller (1959), two examples of a ring NH to ring N hydrogen bond are given: xanthazol monohydrate (Nowaeki & Bfircki, 1955) with a distance of 2.88/l , and isocyanic acid (Senko & Templeton, 1958) with N - H - - - N = 2-90 A. These distances are rather short. Although no attempt was made to locate the hydrogen atoms in these two compounds, disorder of their positions

K A R S T H O O G S T E E N

T a b l e 8. Princ ipal axes and direction cosines with respect to a * , b, and c of thermal ellipsoids

Axis gi(1) gi(2) gi(3) Bila N ( 1 ) 1 0 1 0 1.225

2 0.435 0 0.900 1.167 3 0.900 0 -- 0.435 0.842

c(2) 1 o 1 o 1.18o 2 --0.178 0 0.984 1.138 3 0.984 0 0.178 0.778

N(3) 1 0 1 0 1.219 2 --0.587 0 0.810 0.882 3 0.810 0 0.587 0.734

C(4) 1 0 1 0 1.119 2 --0.273 0 0.962 1.051 3 0.962 0 0.273 0.794

C(5) 1 0 1 0 1.243 2 - -0 .499 0 0.866 1.076 3 0.866 0 0.499 0.879

c(6) 1 o 1 o 1.255 2 0.962 0 --0.273 1.046 3 0.273 0 0.962 0.911

C(7) 1 0 1 0 1.713 2 0.459 0 0.888 1.544 3 0.888 0 --0.459 0.807

0(8) 1 0 1 0 1.904 2 -- 0.204 0 0-979 1.520 3 0.979 0 0.204 0.768

0(9) 1 0 1 0 1.752 2 0.534 0 0.845 0.885 3 0.845 0 --0.531 0.869

C(10) 1 0 1 0 1.690 2 0.812 0 - -0 .584 1.221 3 0.584 0 0.812 0.735

N ( l ' ) 1 0 1 0 1-252 2 0.872 0 0.488 1.115 3 - -0 .488 0 0.873 0.975

Axis gi(1) g~(2) g~(3) Bila C(2') 1 0 1 0 1.365

2 0"999 0 0.037 1.204 3 --0.037 0 0.999 0.926

N(3') 1 0 1 0 1.339 2 0.919 0 - -0 .394 1.196 3 0.394 0 0-919 0.872

C(4') 1 - -0 .707 0 0.707 1.096 2 0 1 0 0.880 3 0.707 0 0.707 0.883

C(5') 1 0 1 0 1.113 2 --0.703 0 0.711 1.078 3 0-711 0 0.703 0.878

C(6') 1 0.855 0 - 0.518 1.055 2 0 1 0 1-013 3 0.518 0 0.855 0.941

N(7") 1 0 1 0 1.436 2 0.272 0 0.962 0.953 3 0.962 0 - 0 . 2 7 2 0.916

C(8') 1 0 1 0 1.348 2 - 0 . 2 1 5 0 0.976 1.184 3 0-976 0 0.215 0.971

N(9') 1 - 0.509 0 0.864 1.255 2 0 1 0 1.198 3 0.861 0 0.509 0.802

:N(10") 1 0 1 0 1.997 2 - - 0.432 0 0.902 1.060 3 0-902 0 0.432 0.755

C(11') 1 0 1 0 1.817 2 --0.543 0 0.840 1.645 3 0.840 0 0.543 0-741

915

along the N - H . . . N hydrogen bond was considered to be possible. The peak attributed to H(2) in the final difference map of the AT base pair is elongated in the direction of N(7') and there is an isolated maximum of 0.30 e.A-8 1.00 A from N(7') in a position where the hydrogen could be expected, if attached to this atom. This, however, cannot be considered as significant evidence for a partial proton transfer along the : N H . - . N bond because spurious peaks with similar heights appear elsewhere in the difference map around other atoms.

I t is also of interest to note the deviation from linearity of the atoms A - H . . . B involved in the hydrogen bonds, as indicated in Fig. 5. None of the deviations from linearity of the atoms involved in the hydrogen bonding between the two bases is exceptionally large. Donohue & Trueblood (1960) observed from a study of molecular models that for N - H - . . O linear, a deviation of about 25 ° for N-H • • • N could be expected. Apparently the depar- tures from the assumed dimensions cooperate in making the experimentally found hydrogen bonding system quite satisfactory. The only other hydrogen

bond present in the structure is the one between the NH2 group of 9-methyladenine and 0(8) of the 1-methylthymine molecule, one unit cell up along the direction of the a axis. Its length is 2-881 /~, and there is no large departure from linearity in this hydrogen bond.

~o--...H~N ,H

/ " - ~ ~o ,,°-=\ i o , " N %

(o) Thymine - AUemne

H H H i ,.... ° I #

H\ iN H % C I N ~ C I N ~ H c~c I I

1 % d \ (b) Cy?o$1ne- Guan ine

H H H . . . . . . . ~ ~,,C~N.,,.C/N H O~C/NI C/~ , .H~N~.C,./N,~c,H

t~%c~C\N. . . / C ~ c / ' N H~C \ C ~ C __C~ / ' " I '"'-"k I \ II \ ~ - ~ . . . . . - /

/ = =~ \ 7 - ~ \ . _ . ~- . i

(¢) Adenine-Guanine (d) Guomne-Adenine

Fig. 6. Some h y d r o g e n - b o n d e d base pairs. F o r exp l ana t i on see tex t .

916 C O M P L E X B E T W E E N 1 - M E T t I Y L T H Y M I N E AND 9 - M E T t t Y L A D E N I N E

In Fig. 6(a) is reproduced the adenine-thymine pair as proposed by Watson & Crick (1953). In this model O(9) and iN(3) of thymine are hydrogen-bonded to the amino group and to N(1) respectively, of adenine. I t is not obvious why in the crystal structure N(3) of thymine is hydrogen-bonded to N(7) and not to N(1) of adenine. Also, the question whether the base pair occurring in the crystal is present in DNA remains unanswered here. We have found it possible to build a DNA molecule from the atomic models of Corey & Pauling (1953) tha t is similar to the double helix proposed by Watson & Crick, but in which the base pairs are hydrogen-bonded in the configuration occurring in the AT crystals. The distance between the two deoxyribose carbon atoms tha t are at tached to the base pairs in this model is 8.645 J~, much shorter than the corresponding distance of about 11.1 /~ i n the model proposed by Watson & Crick. The two deoxyribose-phosphate chains at the outside of the molecule have, as a result of this, al ternatingly smaller and larger vertical separation than they have in the Watson-Crick model.

The arrangement of a guanine-cytosine pair tha t has the same distance between the deoxyribose carbon atoms at tached to the base pairs as in the adenine- thymine pair is shown in Fig. 6(b). In the tautomeric forms tha t are accepted for these molecules neither N(3) of cytosine nor N(7) of guanine has a hydrogen atom available for a hydrogen bond between these two atoms. If the configuration occurring in the crystal structure of AT is also present in DNA, there should be added a proton to N(3) of cytosine or to N(7) of guanine in the base pairs of these two molecules. This would be possible if, for example, one of the corresponding nucleotides is incorporated in a zwitter- ion configuration. The question as to which one might be a zwitterion can presumably be answered on the basis of the assumption tha t it is primarily the dimensions of the adenine-thymine and guanine- cytosine base pairs which exclude the occurrence of other pairs in the DNA molecule. If deoxyguanylic acid is incorporated in the zwitterion configuration with the proton residing on N(7), the formation of an adenine-guanine pair as shown in Fig. 6(c) would be possible, and the dimensions are such tha t this pair could replace the adenine-thymine or guanine- cytosine pairs with the configuration occurring in the AT crystals. In this case the specificity of the base pairing would be lost. If, however, deoxycytidilic acid is present in the zwitterion form, no other pairs

can be formed and the specificity is preserved. In this configuration, cytidine would have two donors, guanine two acceptors, adenine and thymidine one donor and one hydrogen acceptor each. Donohue & Trueblood (1960) have already pointed out tha t i t is possible to form a base pair of adenine and guanine as shown in Fig. 6(d). The dimensions of this base pair are such tha t it presumably could be inserted into the double helix proposed by Watson & Crick, again destroying the specificity.

Numerous a t tempts have been made to grow crys- tals containing other combinations of purine and pyrimidine derivatives. Aqueous solutions containing equimolar quantities of all possible pairs of the four compounds: 1-methylthymine, 9-methyladenine, 1- methylcytosine, and 9-methylguanine have so far failed to yield crystals containing hydrogen-bonded pairs other than the thymine-adenine pair. In all other cases only crystals of the separate compounds could be detected.

I should like to thank Prof. R. B. Corey and Dr R. E. Marsh for their continuous interest during this investigation and the reading of the manuscript.

R e f e r e n c e s

BERGHUIS, J., HAANA~PEL, IJ. ~¢[., POTTERS, M., LOOP- STRA, B. 0., MACGIIJ~AV-RY, C.H., & VEEhrENDA.AL, A.L. (1955). Acta Cryst. 8, 478.

BOOTH, A. D. (1948). Fourier Technique in X-ray Organic Structure Analysis. Cambridge University Press.

BROOMHEAD, J. M. (1948). Acta Cryst. 1, 324. CocBmAN, W. (1951). Acta Cryst. 4, 81. COREY, R. B. t~ PAIILING, L. (1953). Rev. Sci. Instrum.

24, 621. DONOnX~E, J. & T~UEBT.OOD, K. N. (1960). J. Mol. Biol.

2, 363. FUIZ~ER, W. (1959). J. Phys. Chem. 63, 1705. HOOGSTEEN, K. (1959). Acta Cryst. 12, 822. HOOGSTEEN, K. (1963). Acta Cryst. 16, 28. HUGKES, E. W. (1941). J. Amer. Chem. Soc. 63, 1737. McWEE~Y, 1~. (1951). Acta Cryst. 4, 513. NOWACKI, W. & B~3RCKZ, H. (1955). Z. KristaUogr. 106,

339. PA~J-LING, L. & CONEY, R. B. (1956). Arch. Biochem.

Biophys. 65, 164. SE~KO, M. E. & TEMpT.ETO~, D. H. (1958). Acta Cryst. l l ,

808. WATSOn, J. D. & CRICK, F. H. C. (1953). Nature, .Lond.

171, 737.