The Three-Dimensional Structure of Insulin

The three-dimensional structure of insulin remained recalcitrant in spite of the knowledge of its primary sequence. The early crystals had been found by Scott (1936) to contain zinc which could be replaced by other divalent metals. The zinc atom is not heavy enough to be unambiguously distinguishable. Eventually it proved possible to introduce uranyl acetate and uranyl fluoride into the insulin molecule and to obtain the three-dimensional structure, first at 2.8 A resolution and then at 1.9 A (see Blundell, Dodson, Hodgkin, and Mercola, 1972). 

Blundell, Dodson, Hodgkin and Mercola reported the three-dimensional structure of insulin. 

Structure-Activity Correlations. This detailed knowledge of the three-dimensional structure of insulin led to the recognition that its biological activity resides in an area of the molecule rather than in specific amino acid residues, just as dimerization and further association of the molecule also depend on an intact spatial structure. The foregoing concept is corroborated by structural modifications of the hormone. The last three amino acids of the B chain can be removed without a loss of activity, but cleavage of the C-terminal of the A chain (Asn ) results in a total loss of activity. Amino acids can be replaced inside the chains only if such substitution does not change the overall geometry of the molecule. The structure-activity relationships of insulin derivatives are inconsistent and not always comparable. 

The amino acid sequence of insulin does not determine its three-dimensional structure. By catalyzing a disulfide-sulfhydryl exchange, this enzyme speeds up the activation of scrambled ribonuclease because the native form is the most thermodynamically stable. In contrast, the structure of active insulin is not the most thermodynamically stable form. The three-dimensional structure of insulin is determined by the folding of preproinsulin, which is later processed to mature insulin. 

Figure 7. Schematic representation of the three-dimensional structure of insulin based on the X-ray analysis of rhombohedral 2Zn insulin crystals and the proposed conformations, based on model building, of proinsulin, IGF-I, IGF-II, and relaxin. 

The primary structure of P. is elucidated by the standard methods of sequence analysis (see Proteins). The conformation of P. is an important determinant of their biological activity. It is stabilized by peptide and disulfide bonds In P. containing unusual bonds and/or constituents other factors also stabilize the conformation. Although X-ray crystallography (see) was used in the elucidation of the three-dimensional structures of insulin and gramicidin S, the conformational analysis of P. is now by preference carried out in solution, using the spectroscopic methods ORD, CD, IR, NMR and ESR. 

The formation of three disulfide bonds in bovine insulin brings parts of the chain that are distant in terms of amino acid sequence into close proximity in the three-dimensional structure of the protein. 

Now we can ask what is likely to happen to the three-dimensional structure of a protein if we make a conservative replacement of one amino acid for another in the primary structnre. A conservative replacement involves, for example, substitution of one nonpolar amino acid for another, or replacement of one charged amino acid for another. Intnitively, one would expect that conservative replacements would have rather little effect on three-dimensional protein structure. If an isoleucine is replaced by a valine or leucine, the structnral modification is modest. The side chains of all of these amino acids are hydrophobic and will be content to sit in the molecnlar interior. This expectation is borne out in practice. We have noted earlier that there are many different molecnles of cytochrome c in nature, all of which serve the same basic function and all of which have similar three-dimensional structnres. We have also noted the species specificity of insulins among mammalian species. Here too we find a number of conservative changes in the primary structure of the hormone. Although there are exceptions, as a general rule conservative changes in the primary structnre of proteins are consistent with maintenance of the three-dimensional structures of proteins and the associated biological functions. 

I The most important properties of a protein are deter-f mined by the sequence of amino acids in the polypeptide chain. This sequence is called the primary structure of the protein. We know the sequences for thousands of peptides and proteins, largely through the use of methods developed in Fred Sanger s laboratory and first used to determine the sequence of the peptide hormone insulin in 1953. Knowledge of the amino acid sequence is extremely useful in a number of ways (1) it permits comparisons between normal and mutant proteins (see chapter 5) (2) it permits comparisons between comparable proteins in different species and thereby has been instrumental in positioning different organisms on the evolutionary tree (see fig. 1.24) (3) finally and most important, it is a vital piece of information for determining the three-dimensional structure of the protein. 

Relaxin is another peptide that can be extracted from the ovary. The three-dimensional structure of relaxin is related to that of growth-promoting peptides and is similar to that of insulin. Although the amino acid sequence differs from that of insulin, this hormone, like insulin, consists of two chains linked by disulfide bonds, cleaved from a prohormone. It is found in the ovary, placenta, uterus, and blood. Relaxin synthesis has been demonstrated in luteinized granulosa cells of the corpus luteum. 

Figure 26-2 Three-dimensional structure of insulin. Residues In chain A are biue, those in B green. The disuifide bridges are indicated in red. (After Biochemistry, 6th ed., by Jeremy M. Berg, John L. Tymoczko, and Lubert Stryer.

Using this basic methodology, the three-dimensional conformation of insulin was determined, and a significant amount of structure-activity information was gained by the study of insulin analogues and insulins purified from different sources. 

Figure 11.5 Three-dimensional structure of the engineered fast-acting insulin, Insulin lispro. Structural details courtesy of the Protein Data Bank, http //www.rcsb.org/pdb/...

In 1933 after a brief stint at Cambridge and Oxford, she returned to Somerville and Oxford in 1934 and remained there for most of her life teaching chemistry. In 1934 she crystallized and X-ray photographed insulin, only the second protein to be studied. She went on to map the molecular structure of penicillin (1947) and vitamin B12 (1956). In the late 1960s, she created a three-dimensional map of insulin. 

Insulin is the principal drug used to prevent ketosis and sustain life in the treatment of patients with type I (insulin-dependent) diabetes mellitus. Delivery of proteins such as insulin is a challenge because of the molecular size and the sensitivity of the molecule to the loss of its biological activity through minor alterations in the three-dimensional structure. The normal mode of delivery of insulin to patients at the present time is through intramuscular, subcutaneous, or intravenous injections. These delivery methods are not ideal because of (1) the need for training of the patient or the caretaker in the basic steps of injection, (2) the fear of needles by patients, and (3) the feeling of pain and possible fibrotic formation at the injection site. These inconveniences could lead to noncompliance. A variety of approaches for insulin delivery have been... 

The primary structures of IGF-I and IGF-II were determined by Rinderknecht and Humbel (1978) and were shown to have 49 and 47% homology, respectively, with human insulin A and B chains (Table I). Both molecules are single-chain polypeptides with three disulfide bridges corresponding to the B chain and an extended A chain of insulin plus a C peptide of 12 or 8 residues. From the known crystal structure of insulin it has been possible to postulate three-dimensional structures for both IGF-I and IGF-II (Blundell et al, 1978). The proposed conformation of IGF-I (Fig. 5) allows an arrange-... 

She received her doctorate, married (her husband was an historian), and was appointed university lecturer and demonstrator at Oxford. She worked on the X-ray determination of three-dimensional structures of biological molecules and in 1949 published a structure for penicillin. The structure of vitamin B 2 required collecting data from several researchers over a 6-year period and the then-novel use of punch card computers. The structure of insulin was determined after 30 years of research. 

---