Figure 11

When a promising hypothesis doesn’t quite work, scientists often attempt to improve the quality of the available experimental data, since superior information can reveal previously indiscernible pointers. In this spirit, Pauling convinced Robert Corey to embark on a long-term project intended to determine the structure of some simple peptides and amino acids—the building blocks of proteins—using X-ray crystallography. Corey plunged into this study wholeheartedly, and by 1948, he and his collaborators at Caltech managed to unearth the exact architecture of about a dozen such compounds. Realizing that all of Corey’s findings about the chemical bond lengths and the angles between different parts of the molecules, as well as the planarity (the atoms lying in the same plane) of the peptide group, agreed precisely with his own previous formulations, Pauling decided to revisit the problem of the structure of the alpha keratin protein. In an account dictated on his (by then ancient) Dictaphone in 1982, Pauling recalled the circumstances:

In the spring of 1948, I was in Oxford, England, serving as George Eastman Professor for the year and as a fellow of Balliol College. I caught cold and was required to stay in bed for about three days. After two days I had got tired of reading detective stories and science fiction, and I began thinking about the structure of proteins.

Pauling started his new onslaught on the puzzler with the assumption that all the amino acids in the alpha keratin should be in a structurally similar position with respect to the polypeptide chain. While still in bed, he asked his wife, Ava Helen, to bring him a pencil, a ruler, and a piece of paper. Keeping each peptide group in the plane of the paper, using heavier and lighter lines to indicate the three-dimensional relationships, and rotating around the two single bonds to the carbon atoms (with the angle of rotation being the same from one peptide group to the next), Pauling created a helix, a spiral-staircase-like structure, in which the polypeptide backbone formed the core of the helix, and the amino acids projected outward (figure 12). To stabilize the construction, Pauling formed hydrogen bonds between one turn of the helix and the next turn, parallel to the helix axis. (Figure 12; a hydrogen bond is a chemical bond in which a hydrogen atom of one molecule is attracted to an atom of another molecule.) He actually found two structures that could work, one of which he called the alpha-helix, and the other the gamma-helix. That Pauling was able to find solutions to the problem with such primitive tools attests to how crucial his previous discovery of the planarity of the peptide group had been. (Figure 11 represents his attempt to reconstruct the original piece of paper from 1948.) Without it, the number of possible conformations would have been much larger. Excited, Pauling asked his wife to bring him a slide rule (long obsolete, this was the most commonly used calculation tool at the time), so that he could calculate the repeat distance along the fiber axis. He discovered that the structure of the alpha-helix was repeating after 18 amino acids in five turns. That is, the alpha-helix had 3.6 amino acids per turn. Alas, to his disappointment, the calculated distance between turns was 5.4 angstroms, and not the 5.1 angstroms hinted at by the X-ray diffraction patterns. The gamma-helix had a hole down its center that was too small to be occupied by other molecules, so Pauling concentrated his attention on the alpha-helix. Feeling fairly confident in the correctness of his solution, Pauling tried very hard to find some way to adjust either the bond lengths or the bond angles so as to decrease the calculated distance from 5.4 to 5.1 angstroms, but he failed to do so. Consequently, even though he was extremely pleased with his alpha-helix, he decided to refrain from publishing the model until he could understand better the reason for the discrepancy in the spacing.