Insulin research in China and the U.K.

Insulin has been extensively studied since it was discovered by Banting and Best in 1921. Early in 1934, Dorothy Crowfoot and John Desmond Bernal obtained the first X-ray diffraction photograph of an enzyme protein: pepsin. In 1935, they took another photograph of a protein hormone: insulin. The chemical structure of protein was unknown until the amino acid sequence of bovine insulin was solved by Fred Sanger and colleagues in 1955. In 1958, the chemical synthesis of bovine insulin started in China through a nationwide collaboration of three institutions: the Institute of Biochemistry in Shanghai, the Institute of Organic Chemistry in Shanghai and Beijing University. The total synthesis of bovine insulin in crystalline form was accomplished in 1965. The success of the synthesis of the first protein in vitro greatly encouraged young researchers in China. Not long afterwards, the project of structural analysis of insulin crystal was carried out in China through the collaboration of the Institute of Biophysics, the Institute of Physics and Beijing University, and succeeded in 1971. In Dorothy Hodgkin's laboratory in Oxford, X-ray diffraction studies of insulin crystals were resumed after about 30 years, and the structure of rhombohedral insulin crystal was solved in 1969. Through insulin research, the Institute of Biophysics in Beijing and the Institute of Biochemistry in Shanghai established scientific collaboration and personal friendship with Dorothy Hodgkin's laboratory in Oxford, and later Guy Dodson's laboratory in York and Tom Blundell's laboratory in London. In 1975, Dorothy Hodgkin wrote a short note, ‘Chinese work on insulin’ in Nature, anticipating closer scientific exchange between the East and the West. In 1982, a bilateral meeting between the Biochemical Societies in the U.K. and China was held in Oxford. Now, the second bilateral meeting held in Shanghai will further promote the collaboration between our two countries.

The determination of crystal size and disorder from the X-ray diffraction photograph of polymer fibres. 2. Modelling intensity profiles

The intensity profile of the X-ray reflection from a crystalline material is related to the lattice disorder and the distribution of crystal sizes through its Fourier cosine coefficients. However, existing methods of obtaining these structural parameters from the coefficients require more than one order of reflection and this is seldom available with polymer fibres. They also rely heavily on the low-order harmonics which are those determined with least accuracy. The development and testing of a method which overcomes this weakness and which is suitable for use with a single order is described. The coefficients are calculated for a model with paracrystalline disorder and an assumed distribution of crystal sizes and the parameters describing this model are refined to minimize the discrepancy between the calculated and experimental values of the coefficients. Provided the distribution of lengths is asymmetric this discrepancy is no greater than would be expected from experimental error and so the assumed model cannot be rejected on the evidence available. Since a range of model parameters all gave equally good agreement with experiment, it was not possible with a single order to obtain a well defined set of values. Diffraction patterns displaying two orders had been chosen and results from the second order were consistent with the first, only a narrow range satisfying both simultaneously. The method was further developed by calculating the intensity profile from the harmonics and using this in the refinement. There was no advantage over using harmonics; indeed, on occasions the refinement algorithm was unstable producing unreliable results.

A computer-based method of measuring the integrated intensities of the reflections on the X-ray diffraction photograph of an oriented crystalline polymer

The method is designed to be used with a batch-processing computer system and will determine the integrated intensities of the spots on an X-ray diffraction photograph of an oriented fibre of a partially crystalline synthetic polymer. It is necessary to assume that the spot boundary is elliptical, that the intensity distribution along any line through the centre of this ellipse is Gaussian, and that the background intensity variation is linear over the region of a spot; these are justified experimentally, although, in the radial direction, the choice of a Gaussian intensity distribution is probably theoretically unsound. The computational procedures correct for minor differences between users in the choice of input parameters and reject bad choices. The method was applied to determine the intensities of the 30 visible spots in the diffraction photograph of oriented poly(trimethyleneterephthalate) which were used in a subsequent structure refinement. successful integrations were obtained for 22 spots, the failures being (1) pairs of similar intensity just resolved by eye, (2) better resolved pairs of which one member is stronger than the other, or (3) very weak. Statistical tests indicated very much better internal consistency of data than is usually obtained with these materials, and enabled a rational weighting scheme to be used in the structure refinement. The R factor of 7.9% obtained is unusually low, indicating much improved accuracy over earlier methods.

Correlation of Crystal Morphology and Structure Using a Modified X-Ray Film Technique

A convenient technique for the determination of the orientation of a solid-state molecule with respect to its crystalline morphology is presented. The process is especially suitable for spectroscopists wishing to orient crystals in preparation for polarized spectroscopy. Correlation of reciprocal axes with the crystal faces is accomplished by pinholes through a 0-level Weissenberg x-ray diffraction photograph, which allows assignment of spindle angles to lateral positions on the film. The correlation is extended to ORTEP drawings where a scaled-down crystal morphology is included in a standard crystal structure plot, enabling a clear picture of molecular orientations within a crystal. Viewing normal to actual crystal faces allows determination of molecular projections onto desired faces for alignment prior to polarized spectroscopy.

Background evaluation and spot recognition on crystallographic diffraction photographs

A computer method is described which provides a means for determination of both background and spot characteristics without the knowledge a priori of the positions of those spots in a crystallographic diffraction photograph. The background is first determined in any arbitrary region by means of a histogram technique. Then the sets of points emerging from the background's statistical tolerances are collected together with simple neighbouring rules. The problem of large areas which do not fit in a computer memory is discussed and solved with an overlapping window technique. The accuracy of the method is checked by means of a simulation (characteristic retrieval of computer generated spots); the precision appears to be better than the classical rectangular box method except for weak spots.

Scattering of X-Rays by Parallel and Antiparallel Layers in Disordered Stacking and Statistical Layer Shift for a-Chitin

Statistical Layer Shift, α-Chitin The intensities scattered by antiparallel/parallel layers were calculated. Difference synthesis maps allowing for statistical layer shifts were computed for a-chitin. The results support the view that there may be antiparallel sequence with each layer displaced at random by +e or -e from a halving position in the ϰ-direction. This is supported by greater breadth of certain reflection of the type lkl compared to 0kl on the X-ray diffraction photograph. This gives strong evidence for some imperfection in the ϰ-direction and the proposed indeterminacy + e is one possible way in which this might arise.