Dorothy Crowfoot Hodgkin

Insulin: The Structure in the Crystal and its Reflection in Chemistry and Biology by

Publisher Summary This chapter reviews the physical, chemical, and biological properties of insulin in the light of the atomic arrangement found in insulin crystals. It also describes the relation of the three-dimensional arrangement of the atoms in the molecule of 2-zinc insulin crystal to the solution properties of insulin (particularly its states of aggregation), to the chemical reaction and chemical modification of the molecule, and to its primary biological activity. Normally the insulin crystals contain two zinc ions to every six molecules of insulin—a hexamer. The slow solution of the crystals provides a method of delaying the action of insulin that closely parallels the methods adopted in the pancreas itself for the storage and release of insulin. Within many β granules, grains can be seen that almost certainly contain zinc insulin hexamers packed in a crystalline array, and in experimental animals diabetes has been induced by chelating agents, such as EDTA, perhaps simply by interfering with normal insulin storage. It, therefore, seems plausible that ready crystallization of insulin in the presence of zinc is a reflection of the storage processes in the β cell.

Transmission of conformational change in insulin

Crystal structures of insulin contain molecules that are similar but not identical in conformation. Packed helices move relative to each other, these shifts being accommodated by motions of side-chain atoms arising from small changes in torsion angles. Such low-energy conformational adjustments can accommodate shifts of no more than ∼1.5 Å. This limits the extent to which conformational changes can be dissipated locally, causing their transmission over long distances.

The crystal structure of insulin: II. An investigation of rhombohedral zinc insulin crystals and a report of other crystalline forms†

X-Ray diffraction photographs have been taken of zinc-free cubic insulin and of monoclinic zinc insulin crystals, prepared from buffers containing phenol. The wet cubic crystals are very small rhombic dodecahedra, a = 76 A, ρ = 1·09 approx. and space group, I23 or I213; there is one insulin monomer (mol. wt about 6000) in the asymmetric unit. Wet monoclinic crystals have a = 62·3, b = 61·8, c = 47·8 A, β = 110·7°, ρ = 1·19 g cm−3 and air-dried crystals have a = 61·4, b = 53·4, c = 43·8 A, β = 134·5°, ρ = 1·29 g cm−3; the space group is P21 and there are six insulin monomers in the asymmetric unit. A more detailed investigation of two rhombohedral forms with space group R3, has been carried out. Wet 2 Zn insulin crystals, prepared from pig insulin in citrate buffers, have a = 82·5 and c = 34·0 A for the hexagonal unit cell, and ρ = 1·24 g cm−3, implying two insulin monomers per asymmetric unit or six monomers and two zinc atoms per rhombohedral cell. This 2 Zn insulin is essentially the same as rhombohedral zinc insulin formerly prepared from cattle insulin in phosphate buffers and studied by X-ray methods. 4 Zn insulin was prepared from citrate buffers similar except for the addition of sodium chloride in place of acetone. Wet crystals have a = 80·7, c = 37·6 A, ρ = 1·245 g cm−3, and contain six insulin monomers and four zinc atoms per rhombohedral cell. Three-dimensional X-ray intensity data have been collected and Patterson distributions calculated. These show that the molecular arrangement in 2 Zn insulin and 4 Zn insulin is very similar; the difference is of the same order of magnitude as that between wet and air-dried crystals.

The Structure of Vitamin BFormula I. An Outline of the Crystallographic Investigation of Vitamin BFormula

The structure of vitamin B12 has been examined by the X-ray analysis of four different crystal structures, air-dried and wet vitamin B 12, the selenocyanate derivative of B12, and a hexacarboxylic acid prepared by the degradation of the vitamin by Cannon, Johnson and Todd. The analysis turned on the possibility of identifying some of the atomic positions in the molecule in very confused appproximate electron density distributions calculated using only evidence of cobalt or cobalt and selenium contributions to phase the terms employed in the calculation. The recognition of atomic sites depended on a general knowledge of atomic sizes and geometry, on chemical evidence of the existence of a nucleotide-like group in the B12 crystals, and on the presence, immediately surrounding each cobalt atom, of a nucleus which was identical in very different crystals. Once the positions of the atoms in the nucleus and nucleotide were identified, the refinement of the electron density distributions in the different crystals proceeded in a fairly direct manner to establish the arrangement of all the remaining atoms in the molecules. The atomic arrangements found lead to the deduction of the chemical structure of a large part of the vitamin B12 molecule and of the stereochemical organization of the whole. Of particular interest is the evident relation of the molecular nucleus found to porphyrins of type I I I and to porphobilinogen.

The crystal structure of insulin: III. Evidence for a 2-fold axis in rhombohedral zinc insulin

The existence of non-space-group symmetry elements in rhombohedral 2 Zn insulin and 4 Zn insulin crystals has been investigated. A rotation function shows the existence of a 2-fold axis perpendicular to the crystallographic c -axis (3-fold) and making an angle of 44° with the a -axis. A translation function, and independent arguments from the Patterson series, show that this is a 2-fold axis, without translation parallel to its length. In 2 Zn insulin the 2-fold axes pass through or within 0·5 A of a 3-fold or, less probably, a 3-fold screw axis; in 4 Zn insulin they are about 1 A from it.

The structure of Vitamin B12 VI. The structure of crystals of vitamin B12 grown from and immersed in water

Vitamin B12 crystals have been grown from water and photographed in their mother liquor by means of copper Kα X-radiation. The crystal unit cell dimensions are a = 25·33 Å, b = 22·32 Å and c = 15·92 Å, space group P212121 and the formula for the asymmetric unit probably C63H88O14N14PCo.22H2O. From the intensities of some 2900 X-ray reflexions visually measured in 1949, calculations have been made of the electron density in the crystal through a series of approximations. The calculations lead to positions of all the atoms in the B12 molecule, not counting hydrogen atoms; these positions agree closely with those derived for the air-dried crystals except over certain of the amide side chains which differ in conformational details. In the wet crystals the molecules make contact with one another through hydrogen bonds between the amide groups. Between them is a region occupied by water molecules. Some eight or nine of these have definite positions in the crystal; the remainder appear considerably disordered probably as a result of movement in the crystal. The disordered positions lie in or near channels which run between the B12 molecules continuously throughout the crystal and through which it seems likely that water passes out on drying.

Structure of Ferroverdin

FERROVERDIN, a green iron-containing pigment, was isolated in 1955 by Chain, Tonolo and Carilli1 from an unidentified species of Streptomyces. It was at first assigned the formula C30H24O8N2Fe and the iron was shown by measurements of magnetic susceptibility to be in the ferrous state2. Later the ligand present was proved to be the p-vinyl phenyl ester of 3-nitroso-4-hydroxy-benzoic acid3,4. X-ray crystallographic measurements were undertaken to find the atomic arrangement in this unusual complex; they show, in two different crystal structures, that each iron atom is attached to three nitrosophenyl ligands and that the charge is balanced by sodium ions.

The Structure of Insulin

It is curious to read once again the text of the paper by Dr. Moses Barron1 which was first read by Fred Banting on October 30, 1920, and which stirred him so urgently to attempt the isolation of insulin. He stopped at the library casually on the way home from the medical school and looked at the first paper in Surgery, Gynecology and Obstetrics. Because he had to talk to students on the following day on the function of the pancreas and because its title was “The relation of the islets of Langerhans to Diabetes with special reference to cases of Pancreatic Lithiasis,” he took the journal home with him. The sentences he underlined were “Arnozan and Vaillard ligated the pancreatic ducts in rabbits and found that within twenty-four hours the ducts became dilated, the epithelial cells were desquamated and there were protoplasmic changes in the acinic cells … Ssobolew ligated the ducts in rabbits, cats and dogs. He found a gradual atrophy and sclerosis of the organ and relatively intact islets and no glycosuria.” At 9 p.m. Banting rang up Dr. Tew to say that he had an idea-that it might be possible to isolate the antidiabetic hormone from the islets by first allowing the acinar cells, containing digestive ferments, to atrophy. It was an idea so powerful that it changed his life and the lives of manv others.

The Structure of Vitamin BFormula. II. The Crystal Structure of a Hexacarboxylic Acid Obtained by the Degradation of Vitamin BFormula

A hexacarboxylic acid, obtained by the degradation of vitamin B12 by Cannon, Johnson & Todd in 1953 has been examined by X-ray analytical methods. These lead to a solution of both the crystal and chemical structure of the acid. The crystals are orthorhombic, a = 24·58, b = 15·52, c = 13·32 Å, space group, P212121, n = 4. The asymmetric unit is found to consist essentially of one molecule of hexacarboxylic acid, C46H58O13N6. CoCl, two molecules of water and one of acetone. The hexacarboxylic acid molecule has a central cobalt atom in approximately octahedral co-ordination attached to one cyanide group, one chlorine atom and four nitrogen atoms of the corrin nucleus. The nucleus itself is substituted by acetic and propionic acid groups, a lactam ring and a number of methyl groups. The position of the cobalt atom in the crystal structure was first found from Patterson projections and the remaining atomic positions then derived from a series of calculated approximations to the three-dimensional electron density distribution. For these calculations, phases were derived from structure factors calculated on gradually increasing numbers of selected atomic positions from the stage of ρ1, where only the cobalt atom sites were known, to ρ10 where 73 atoms, not counting hydrogen, had been placed. The process was not quite straightforward; particular difficulty was experienced in finding the positions of the atoms of one side-chain which may be affected by disorder. The parameters of the atoms have been refined by two cycles of least-squares calculations. A number of observations were made in the course of the analysis which bear on the further use of non-centrosymmetric Fourier syntheses in the study of complex structures. An appendix by A. Vos deals with intensity anomalies observed on the X-ray photographs of the hexacarboxylic acid which provide evidence of its absolute configuration. An appendix by K. N. Trueblood summarizes various aspects of the analysis of the hexacarboxylic acid, seen as a whole.

The Structure of the Corrin Nucleus from X-ray Analysis

The name corrin was adopted during the first international symposium on vitamin B12 at Hamburg in 1957 as the name of the nucleus shown in I and considered to be present as a cobalt complex in cyanocobalamin. The name was chosen democratically by a vote of the meeting as a whole from among several proposed alternatives. It had in fact been suggested by Dr Karl Folkers and conformed with the names chosen at that time for other B12 derivatives, since it began with the letters Co. In the case of corrin, this prefix gave rise to some misgivings; all the other derivatives named contained cobalt, corrin only the memory of the first cobalt compounds in which its presence had been deduced.