J. C. KENDREW 



Unfortunately crystals of myoglobin, though relatively easy to pre- 

 pare by salting out the aqueous solution of the protein with ammonium 

 sulphate, are normally obtained in a habit which is most unsatisfactory 

 from the point of view of the x-ray crystallographer, whose ideal is a 

 single crystal with no dimension less than about 0-25 mm. The myo- 

 globin crystals are amply long in one dimension, but extremely small 

 in the other two — in other words they form thin needles and, as 

 ordinarily prepared, lie just beyond the useful range of present-day 

 single crystal x-ray technique. 



The fortunate circumstance which made x-ray work possible was 

 the discovery by M. W. Rees 3 of the Department of Biochemistry at 

 Cambridge that, if myoglobin is crystallized from very concentrated 

 phosphate buffer solutions instead of from ammonium sulphate, it 

 assumes a rather different habit, the needles becoming flattened to long 

 lath-like plates (see Figure 1). Phosphate concentrations between 3 

 and 4 molar, and concentrated protein solutions, are required ; mix- 

 tures of NaH 2 P0 4 , 2H a O and K 2 HP0 4 were used since these are the 

 most soluble acid and basic phosphates available, the proportions 

 being adjusted to give a pH of about 6-4. Details of an almost exactly 

 similar crystallization of whale myoglobin (but at slightly higher pH) 

 are given by J. Keilin and K. Schmid 4 . The crystals so prepared are 

 often several mm. long and about one mm. across, but even the largest 

 of them is still very thin, so that x-ray exposures of up to 200 hours 

 were necessary to obtain usable diffraction pictures. After irradiation 

 of this duration protein crystals appear to suffer some sort of internal 

 disintegration so that, although outwardly unaltered, their diffraction 

 patterns become weaker and weaker. This limits the amount of in- 

 formation which can be obtained from small crystals, and in these 

 myoglobin crystals this limit has just about been reached. This is the 

 reason why it is unlikely that the results about to be described can be 

 greatly added to until even larger crystals can be prepared. 



CRYSTAL SYMMETRY AND CELL DIMENSIONS 



Myoglobin crystals are monoclinic, that is to say the unit cell dimen- 

 sions are all unequal, while the angles between the a and b axes, and 

 between the c and b axes, are 90° each, and the angle /? between a and c 

 has a value different from 90°. The actual dimensions are : a = 57-3 A, 

 b — 30-8 A, c = 57-0 A, fi = 112°, and a perspective drawing of the 

 cell is given in Figure 2a. For comparison a drawing of the crystal in 

 the same orientation is given in Figure 2b. 



The space group is P2 X . This implies that a screw diad axis runs 

 through the cell parallel to b. The meaning of this is that any piece 



150 



