Brilliant Blunders: From Darwin to Einstein - Colossal Mistakes by Great Scientists That Changed Our Understanding of Life and the Universe

by Livio, Mario

  Given this formidable legacy, we can imagine that when Pauling saw the title of the paper by Bragg, Kendrew, and Perutz, his heart missed a beat. The first two paragraphs of the paper indeed gave the impression that Bragg’s team may have beaten him to the punch: “Proteins are built of long chains of amino-acid residues . . . In this paper an attempt is made to glean as much information as possible about the nature of the chain from x-ray studies of crystalline proteins, and to survey the possible types of chain which are consistent with such evidence as is available.” Pauling read quickly all thirty-seven pages and was relieved to discover that while the Cavendish researchers described some twenty structures, the alpha-helix wasn’t one of them. Moreover, they concluded that none of the examined structures was acceptable as a model for alpha keratin. Pauling agreed happily with this conclusion, especially since he thought that Bragg’s team did not apply the most important constraint to its configurations but did impose a handicap that he regarded as totally unnecessary. On one hand, none of Bragg’s models assumed the planarity of the peptide group, of whose correctness Pauling was absolutely convinced. On the other, the Cavendish team appeared to be hung up on the notion that in every full turn of its helical structures, there had to be an integer number of amino acids. Pauling’s alpha-helix broke with tradition and had about 3.6 amino acids per turn, and he saw nothing wrong with that. Coming from an X-ray crystallography background, Bragg also adhered religiously to the apparent 5.1 angstrom distance between turns suggested by Astbury’s data. Perutz later described that to start the team off, Bragg hammered nails representing amino acid residues into a broomstick in a helical pattern with an axial distance between successive turns of 5.1 centimeters.

  Pauling was always extremely competitive in nature. Even though he was pleased to see that the Cambridge team had missed a few key points, the appearance of Bragg’s paper prompted him into action, for fear he might be scooped. In October 1950 he and Corey sent a short note describing the alpha-helix and the gamma-helix to the Journal of the American Chemical Society. Around the same time, some encouraging results were coming from another British research group at Courtaulds Research Laboratories. There, Clement Bamford, Arthur Elliott, and their collaborators succeeded in producing fibers of synthetic polypeptides. To Pauling’s delight, X-ray diffraction photographs of those fibers showed clearly that the repeat distance along the axis was 5.4 angstroms—consistent with Pauling’s findings—rather than 5.1 angstroms. This raised the suspicion that the latter feature in the X-ray photographs of hair could simply be an artifact produced by overlapping reflections rather than a major clue to the structure. Increasingly convinced of the truth of this interpretation, Pauling submitted a paper by himself, Corey, and Branson that contained a detailed explanation of the alpha- and gamma-helices. It was only fitting that this important paper was submitted precisely on the day of Pauling’s fiftieth birthday, February 28, 1951.

  There is, incidentally, an interesting anecdote concerning the use of the term “helix,” which I heard from chemist Jack Dunitz, who at the time was a postdoctoral fellow with Pauling. Dunitz recalled that in 1950 Pauling kept using the term “spiral” to describe the structure of alpha keratin. Even in Pauling and Corey’s short communication in the Journal of the American Chemical Society, they wrote exclusively about spirals. One day, said Dunitz, he remarked to Pauling that he thought that the word “spiral” referred only to the two-dimensional, planar shape, while the three-dimensional one had to be called a “helix.” Pauling responded that a spiral could be either two-dimensional or three-dimensional, but added that on second thought, he liked the word “helix” better. When the extensive manuscript by Pauling, Corey, and Branson was submitted, it avoided the word “spiral” altogether. Its title read: “The Structure of Proteins: Two Hydrogen-Bonded Helical Configurations of the Polypeptide Chain.” Pauling was by then so confident in his model that he and Corey followed the alpha-helix paper with a barrage of papers on the folding of polypeptide chains.

  That spring in England, Max Perutz went one Saturday morning to the library, and there, in the latest issue of the Proceedings of the National Academy of Sciences, he found the series of papers by Pauling. Some thirty-six years later, he described what he had experienced that morning (in a somewhat technical language, but the emotions were crystal clear):

  I was thunderstruck by Pauling and Corey’s paper. In contrast to Kendrew’s and my helices, theirs was free of strain; all the amide groups were planar and every carboxyl group formed a perfect hydrogen bond with an amino group four residues further along the chain. The structure looked dead right. How could I have missed it? Why had I not kept the amide groups planar? Why had I stuck blindly to Astbury’s 5.1 angstrom repeat? On the other hand, how could Pauling and Corey’s helix be right, however nice it looked, if it had the wrong repeat? My mind was in a turmoil. I cycled home to lunch and ate it oblivious of my children’s chatter and unresponsive to my wife’s inquiries as to what the matter was with me today.

  Thinking a bit more about Pauling’s model, Perutz realized that the alpha-helix resembled a helical staircase, in which the amino acid residues (marked by “R” in figure 12) were forming the “steps.” The height of each step was about 1.5 angstroms. Bragg’s X-ray diffraction theory therefore predicted the existence of never-before-reported X-ray reflection signatures, separated by 1.5 angstroms, from planes perpendicular to the fiber axis. None of the models of Bragg’s group would have produced such a mark, while this would have been a distinct “fingerprint” of Pauling’s alpha-helix.

  Just as he was about to conclude that the lack of such reflections in Astbury’s data was sufficient to refute Pauling’s model, Perutz suddenly recalled that Astbury’s particular experimental setup—with the fibers oriented such that their long axes were perpendicular to the beam of X-rays—would not have really allowed for the detection of the 1.5 angstrom signature. Rather, calculations predicted that the optimal conditions to observe the reflection would have required inclining the fibers at an angle of about 31 degrees.

  Perutz felt absolutely compelled to make the crucial test right away. He cycled back to the lab, grabbed a horsehair he had in a drawer, inserted it into the apparatus at the angle he calculated to be favorable for detecting the reflection, put a film around it (as opposed to Astbury’s flat-plate camera, which was too narrow and could have missed reflections deflected at large angles), and fired the X-ray beam. The few hours that passed before he could develop the film were sheer agony, but finally Perutz had the answer. The strong reflection predicted by the alpha-helix at a spacing of 1.5 angstroms stuck out unambiguously!

  Perutz showed the X-ray photograph to Bragg first thing on Monday morning. Bragg wondered what it was that suddenly gave Perutz the idea to conduct this crucial test. Perutz replied that he was madly furious with himself for not having thought of the alpha-helix. Bragg retorted with what has by now become an immortal phrase: “I wish I had made you angry earlier!”

  Life’s Blueprint

  Not everything that Pauling wrote in that famous series of papers from 1951 was correct. A careful scrutiny of his entire oeuvre for that year reveals several weaknesses. In particular, the gamma-helix eventually had to be abandoned. These minor shortcomings, however, don’t take away anything from Pauling’s groundbreaking achievement: the alpha-helix and its prominent role in the structure of proteins. Pauling’s contributions to our understanding of the nature of life were substantial. He was one of the first scientists to see that in spite of its inherent complexity, biology is, at its core, molecular science augmented by the theory of evolution. Already back in 1948, he wrote perceptively: “To understand all these great biological phenomena we need to understand atoms, and the molecules that they form by bonding together; and we must not be satisfied with an understanding of simple molecules . . . We must also learn about the structure of the giant molecules in living organisms.”

  Pauling’s influence on the general theory and methodol
ogy of molecular biology was equally impressive. First, in his seminal 1939 book The Nature of the Chemical Bond and the Structure of Molecules and Crystals: An Introduction to Modern Structural Chemistry, he remarked prophetically on the importance of the hydrogen bond for biomolecules: “I believe that as the methods of structural chemistry are further applied to physiological problems it will be found that the significance of the hydrogen bond for physiology is greater than that of any other single structural feature.” Indeed, the structure of many organic molecules, ranging from proteins to nucleic acids, confirmed this prediction fully.

  Second, Pauling pioneered model building and turned it into a predictive art form based on strict rules of structural chemistry. Even the space-filling colored models developed at Caltech became a hot item in the arena of macromolecular research. These models, produced for labs by the Caltech workshop, fetched as much as $1,220 in 1956 for a set that contained about six hundred atom models.

  Pauling’s practice of using the X-ray diffraction patterns not as the starting point but as the ultimate arbiter among sophisticated, educated guesses also proved to be enormously effective—Watson and Crick were about to apply the same approach to the structure of DNA.

  There was another remarkable observation concerning genetics that Pauling made in a lecture in 1948, but apparently even he did not realize at the time its full implications. In the first part of that lecture, Pauling reminded his audience:

  The Gregorian monk Mendel noted that the inheritance of characters by pea plants, such as the character of tallness or of dwarfness, or the character of having purple flowers or white flowers, could be understood on the basis of hereditary units transmitted from the parent to the offspring. Thomas Hunt Morgan and his collaborators identified these units with genes arranged in a linear array in the chromosomes.

  Then, toward the end of the lecture, he added the following comment:

  The detailed mechanism by means of which a gene or a virus molecule produces replicas of itself is not yet known. In general the use of a gene or virus as a template would lead to the formation of a molecule not with identical structure but with complementary structure. It might happen, of course, that a molecule could be at the same time identical with and complementary to the template on which it is moulded. However, this case seems to me to be too unlikely to be valid in general, except in the following way. If the structure that serves as a template (the gene or virus molecule) consists of, say, two parts, which are themselves complementary in structure, then each of these parts can serve as the mould for the production of a replica of the other part, and the complex of two complementary parts thus can serve as the mould for the production of duplicates of itself [emphasis added].

  As we shall soon see, had Pauling remembered his own pronouncement four years later when he was trying to determine the structure of DNA, he might have avoided making a terrible blunder.

  Pauling started to turn his attention to DNA only in the summer of 1951. Until the early 1950s, most life scientists subscribed to the protein paradigm: the view that proteins, rather than nucleic acids, formed the foundation for life and were the crucial players in reproduction, growth, and regulation. The roots of this view could be traced to biologist Thomas H. Huxley (“Darwin’s Bulldog”), who believed that the protoplasm—the living part of the cell—was the source of all of life’s attributes. Proteins, which are built up of amino acids in a long chain, make up a large fraction of all living cells, while nucleic acids, as their name implies, were found first in the nuclei of cells.

  The early work on the structure and constitution of the nucleic acids by biochemist Phoebus Levene did not help to spark interest in these molecules. If anything, his studies achieved precisely the opposite effect. Levene managed to distinguish the deoxyribonucleic acid (DNA) from ribonucleic acid (RNA), and to find some of their properties. But his results generated the impression that these were rather simple and dull substances unsuited for the complex tasks of governing growth and replication. In the words of cytologist Edmund Beecher Wilson (in 1925): “The nucleic acids of the nucleus are on the whole remarkably uniform . . . In this respect they show a remarkable contrast to the proteins, which, whether simple or compound, seem to be of inexhaustible variety.” This impression persisted throughout the 1940s. By then, DNA was known to be composed of unbranched chains of units called nucleotides. The nucleotides themselves also appeared to be fairly uncomplicated, with each one containing three subunits: a phosphate group (a phosphorus atom bonded to four oxygen atoms), a five-carbon sugar, and one of four nitrogen-containing bases. The four bases were: cytosine and thymine, which were single ringed; and adenine and guanine, which were both double ringed (see figure 13). What was still not known, even in 1951, was the actual structure: how exactly the subunits connected to each other to form nucleotides, and the nature of the links between the nucleotides themselves. However, while all of this seemed to be fairly interesting from a chemical perspective, at the end of 1951 most geneticists still believed that DNA’s only role was structural, acting perhaps as a scaffold for the more sophisticated proteins rather than being directly related to heredity.

  This fact in itself was somewhat surprising, given that in a paper published back in 1944, biologists Oswald Avery, Colin MacLeod, and Maclyn McCarty provided strong experimental evidence that the genetic material of living cells was composed of DNA. Avery and his colleagues grew large quantities of virulent bacteria, and after managing to separate them into their biochemical constituents, they concluded that DNA molecules—and not proteins or fats—were the components responsible for converting nonvirulent bacteria into virulent ones. In a May 1943 letter describing the results to his bacteriologist brother, Roy, Avery concluded, “So there’s the story, Roy—right or wrong it’s been good fun and lots of work.” The reason that Avery’s findings did not get the attention they deserved may have had to do with the fact that since none of the three scientists was a geneticist, their conclusions were formulated with such caution that many of the life scientists failed to appreciate their full import. The statement in the paper read: “If it is ultimately proved beyond reasonable doubt that the transforming activity of the material described is actually an inherent property of the nucleic acid, one must still account on a chemical basis for the biological specificity of its action.” Still, careful readers should have taken notice of the paper’s summary: “The data obtained . . . indicate that, within the limits of the methods, the active fraction contains no demonstrable protein . . . and consists principally, if not solely, of a highly polymerized, viscous form of desoxyribonucleic acid [DNA].”

  Figure 13

  Pauling was familiar with Avery’s work, but even he admitted in a later interview that at the time he did not believe that DNA had much to do with heredity: “I knew the contention that DNA was the hereditary material. But I didn’t accept it; I was so pleased with proteins, you know, that I thought that proteins probably are the hereditary material, rather than nucleic acid.” Chemist Peter Pauling, Linus’s son, also affirmed that this had indeed been his father’s attitude. In a short article written in 1973, Peter reported, “To my father, nucleic acids were interesting chemicals, just as sodium chloride [ordinary table salt] is an interesting chemical, and both presented interesting structural problems.”

  Nevertheless, toward the end of 1951, an unusual paper by biochemist Edward Ronwin, then at the University of California at Berkeley, intrigued Pauling sufficiently to prod him into action. The paper, entitled “A Phospho-tri-anhydride Formula for the Nucleic Acids,” appeared in November 1951. In it, Ronwin proposed a new “design” for DNA, in which each phosphorus atom connected to five oxygen atoms, while Pauling—the consummate structural chemist—was absolutely convinced that it had to link only to four. Annoyed, Pauling fired a quick communication to the editor of the Journal of the American Chemical Society (together with chemist Verner Schomaker) in which they first noted that “in formulating a hypothetical structure for a sub
stance, one must take care that the structural elements of which use is made are reasonable ones.” Their conclusion was even more dismissive: “The ligation of five oxygen atoms about each phosphorus atom is such an unlikely structural feature,” they said, that the proposed formula for DNA “deserves no serious consideration.” Ronwin retorted by pointing out that other substances in which phosphorus was bonded to five oxygen atoms did exist. Pauling and Schomaker had to withdraw their disparaging statement, but they still insisted correctly on the fact that structures of this type were extremely sensitive to moisture, which made them unlikely candidates for DNA. This exchange would have been insignificant except that it did get Pauling thinking about how DNA might be constructed. To make progress, however, he needed high-quality X-ray diffraction photographs of DNA, since the ones available in print were old photos taken by William Astbury and Florence Bell in 1938 and 1939. Unfortunately, good X-ray photos were not easy to come by. Caltech did produce new photographs in the early 1950s, but, surprisingly, those turned out to be of inferior quality to those of Astbury and Bell. While weighing his options, Pauling heard that Maurice Wilkins of King’s College, London, had generated what were described as “good fibre pictures of nucleic acid.” Deciding that he had nothing to lose, Pauling wrote to Wilkins to inquire whether the latter was prepared to share those photos. Unbeknown to Pauling, however, the activity around DNA in England was rapidly approaching frenzy.