A Reader's Guide to The Double Helix
© 2009 Kenneth R. Miller
"Francis: Do you think we were lucky to have solved it?"
[Click for Video]
The Double Helix by James D. Watson is an extraordinary manuscript, one that stands nearly alone in the history of science writing. It is a first person account of one the major scientific discoveries of all time, and for that reason alone it is worth reading. As reviewers and analysts have noted, The Double Helix is also an exceptional piece of writing. It is lively, engaging, and even scandalous in parts, shattering the myth that great science is done in an atmosphere of dispassion and objectivity.
Most parts of The Double Helix are so clearly written that they stand without comment, and this document is not an attempt to review or analyze the book. In a few places, however, a few words of scientific explanation might be useful to the first-time reader, and I have prepared these notes for exactly that purpose. The notes are keyed to specific chapters in the text, and they may help you with portions of the text that may seem a bit confusing. I have included photos which the Cold Spring Harbor Laboratory has made available for educational purposes, and links to Oregon State's page of videos documenting the Race for DNA.
I hope that this Reader's Guide will be freely copied and used for educational purposes. I should also recommend two other books for the reading list of any serious student of these events, especially where the contributions of Rosalind Franklin are concerned:
Watson was able to persuade Sir Lawrence Bragg (right) to write the introduction to his book, which ensured that it would be taken seriously by fellow scientists. Bragg's contributions to the field of structural biology are legendary. In fact, the equation that describes the relationship between the spacing between spots on a diffraction pattern and atoms in a crystal is known as Bragg's Law. It was named for his father, who also won a Nobel prize . . . . but the younger Bragg won his Nobel at the age of 25 (Still the youngest person ever to win a Nobel). It's also worth noting Bragg's advice that the book should be read "in a forgiving spirit." An interesting observation for a book's Foreword, don't you think?
"... I do not believe that the way DNA came out constitutes an odd exception to a scientific world complicated by the contradictory pulls of ambition and the sense of fair play." I agree. And I think that most other scientists would agree, as well, which makes the manuscript that much more interesting. As Watson notes, no two people ever see the same events in exactly the same light something to keep in mind as we approach the book's more controversial sections.
"How's Honest Jim?" ." That's the ringing phrase that Watson relates from a brief encounter with a colleague in the summer of 1955, more than two years after the publication of the double helix manuscript. By the end of the book, it will be apparent that the phrase "Honest Jim" was meant sarcastically.
"The Cavendish" was a dumpy, run-down chemical lab at Cambridge University. The structure of Cambridge University is quite different from that of a US college. There are a number of semi-independent "colleges" bound together, each with its affiliated faculty (known as fellows) who work in a formal tutorial program for students. At designated meals the fellows eat together in the college dining hall at a special table on an elevated platform ("High Table") and are expected to engage in enlightened conversation. The original purpose of high table, I was told during a brief stint of my own at Pembroke College in Cambridge, was to enable the tutors and professors to keep an eye on the students during their meals to prevent misbehavior.
This chapter contains the first of several references to Erwin Schrödinger's book, What is Life, based on a collection of lectures that the great physicist gave during World War II. The book deals with many topics, but to many of those who read it, its principle message was that the mechanism of heredity, indeed of life itself, should be open to investigation by the techniques of physics and chemistry. Schrödinger was certain that great surprises and great rewards were to be found from such investigations.
As an American, I found it amusing that Britain was thought to be small enough that one lab (in this case in London) working on a problem was thought to be sufficient. Watson uses this observation to launch the first of many remarkable slurs on page 19 (page 13 of the Norton Critical Edition), this one directed against Britain's European neighbor ("In France, where fair play obviously did not exist...").
Also note Watson's explanation that Franklin was "Wilkins' assistant." Anne Sayre, in her book on Rosalind Franklin, points out that Franklin was, in fact, not Wilkins' assistant, so it is hardly surprised that she realized this, too. On a purely scientific note, it is quite surprising to learn that Wilkins had refused a request by Pauling to have a look at some diffraction patterns (p. 21; p. 15, Norton edition). It's very unusual, even today, for any scientist to refuse another a look at their data. On a related subject, the last sentence of this chapter, so outrageous that it may make your blood boil, sets the tone for Watson's treatment of Rosalind Franklin. "...the best home for a feminist was in another person's lab." By this point in the book, the personal tone of his narrative, for better and for worse, is already set. Watson is going to say what he thinks, no matter how others will react.
Max Delbrück's name comes up in this chapter. Delbrück, who helped to pioneer the use of bacteriophage (bacterial viruses) as an experimental system, was one of the most influential of a group of scientists who moved from physics into biology near the end of the 1940s.
The backdrop to the discussion in this chapter is the fact that experiments by Oswald Avery (right) at Rockerfeller in New York had established (for the first time) that DNA was the genetic material. Not everyone yet believed these results, but at this point everyone who would matter to Watson had already concluded that genes were made from DNA, rather than proteins (as many people had previously believed). Watson, however, had already decided that the structure DNA was the question to study, even if his fellowship required him to learn biochemistry instead.
It's wonderful to read Watson's openness about his ignorance of X-ray crystallographic techniques. In fact, to me one of the great charms of this book is its author's remarkable openness about his own flaws and mistakes, as well as his blatant recollection of his own prejudices. The truth is that very few people outside of the X-ray community (even today) really understand the technique, and that in scientific meetings where X-ray data are presented, the crystallographers gloss over the details of the technique and get right to the structures, which are what most people are really interested in.
The focus of this chapter is on personality rather than on science, and that points out an interesting (and probably pretty obvious) truth. The process of science, from fellowships to collaborations, hinges as much on personalities, on conflict and on friendship, as it does on experimentation. This chapter also contains an interesting secondhand account of Linus Pauling's first European lecture on the alpha-helix structure of proteins (photo of Pauling with the alpha-helix model is at right). Pauling's achievement was later to win him the Nobel prize, and at the time of the lecture the British crystallographers clearly regarded him with a mixture of awe and resentment. Watson seems to have regarded Pauling's success with proteins as the very model of what he hoped to do with DNA, even to the point of composing (in his head) a line with which to open a paper describing his success.
This chapter describes the way in which a funding agency turned down Watson's request to switch the subject of his fellowship to DNA, which, if he had gone along with it, might have ended the "story" of this book right there. Promising a funding agency that you will do one thing (work with Roy Markham) and then going into another lab to do something quite different is a well-established scientific tradition. Watson enjoyed this tradition as much as anyone.
Take note of the importance of model-building in Pauling's work on proteins. The "trick" that Watson attributes to Pauling's successes with proteins involved not waiting until every aspect of an X-ray diffraction pattern could be rigorously accounted for. Rather, Pauling had build paper and wire models of the amino acid chains in proteins (polypeptides, actually) and twisted them into a variety of possible shapes. He then compared this models with the actual data to see if any of them came close to accounting for the X-ray patterns. It was from such work that the alpha-helical pattern of protein folding was solved. Watson had obviously decided early on that he would have a better chance at success if he followed Pauling's example of model-building rather than experimentation. But even a model-builder needs data! And the only data available seemed to be the most minimal chemical information on the nature of the polynucleotide chains that make up DNA (Figure on page 40; p. 33 Norton).
One of the key technical aspects of this book are the repeated discussions of how X-rays might be diffracted by a helical molecule. Remember that a helix may be viewed from many different directions. Viewed down the long axis, a helical structure looks very much like a tube or barrel, and there should be no clues in the diffraction pattern that it is fact is a helix. Viewed from the side, however, a helix should zig-zag back and forth, and this could produce a characteristic pattern in a good X-ray picture, like the diagram below.
Note that the pattern is in the form of an "X," which will appear later in Franklin's superb diffraction patterns of the "B-form" of DNA. An appreciation of this was the essence of Crick's solution to the diffraction of helical molecules, which is described at the end of this chapter.
This very subjective recounting of a seminar on DNA by Franklin (at right) is an important part of the book, because it illustrates how critical her thinking was in developing the double helix model. As Watson makes clear, Franklin was not a fan of the use of molecular models to solve structures, which is quite true. He regards her reluctance to engage in model-building as a negative, but it also helps to account for her insistence on getting better and better data, which would be the key to ultimate success in the race to solve the structure of DNA.
Watson minimizes the importance of this talk, but he also mentions her measurement of the water content (p. 52; or p. 46) of the DNA samples. This was a key element in getting good diffraction pictures of DNA. Remember that living cells are mostly water, and therefore DNA interacts with water all of the time. Franklin suspected, correctly as it turned out, that DNA samples would have to have a high water content in order to have the same structure that they did in living cells. If too much water is taken out in an effort to make the samples crystalline, DNA interacts with itself and the structure changes. This was a point that Watson and Crick were not to appreciate until the very end of their collaboration on DNA, but it is something that Franklin knew from the very beginning of the story.
Note (in case you are reading this book with an eye towards Watson's sexism) that he offhandedly mentions that Dorothy Hodgkin was the best crystallographer in England. Once again, Watson's remarkable candor allows him to make that statement and also to confess that he was the "wrong person" to listen to Franklin's seminar the previous day. He clearly did not understand it, and in this chapter he tells us so.
Also in this chapter, Watson and Crick wonder about how many strands might be in the helix. Pauling's alpha-helix for proteins had just one strand. But we've already learned that the diameter of DNA (about 20 angstroms) was wider than one would expect for a single polynucleotide strand. The number of strands could only be determined by knowing the angles at which the helix appeared to zig-zag. If there were, say, 4 strands, then the individual strands should be nearly parallel to the long axis of the fiber. If there were only 2, the angle should be much sharper. That's why getting a good X-ray pattern that showed the angle clearly would be of paramount importance. And Franklin's X-ray patterns were beginning to look clear and regular enough to promise that such data would be forthcoming.
Note in my rough sketches (at right) how the bending angle is much sharper in the 2-fiber helix than it is in a 3-fiber helix of the same diameter:
The change in a bending angle between a 2-helix model and a 3-helix model would affect the angle of that "X" in the diffraction pattern, meaning that the angle measured from the pattern could be the crucial clue as to how many strands were in the molecule.
Note on page 56 (Norton p. 51) that Watson & Crick had decided that the sugar-phosphate backbone had to be near the center of the fiber in order to ensure regularity of structure (since the bases were so different from each other in terms of size and chemistry). But they also realized that there was a problem, namely, how to handle the negatively charged phosphate ions, which should have repelled each other, making such a structure impossible! Clearly, this was a difficult problem one that was to lead to their first, disastrously incorrect, attempt to solve the riddles of DNA. Watson gives us a sarcastic hint of this when he refers to "our impending triumph with DNA" on page 58.
The figure on page 60 (Norton p. 54) is extremely important. The fact that the plane of a base and the plane of the deoxyribose sugar are at right angles means that the bases are normal to the long axis of a polynucleotide strand, like my sketch at right:
This was a crucial insight to building a 3D model. Without knowing the planes of the bases, the model-builder could not be sure how to position the major chemical groups of each nucleotide.
However - if the molecule was a multistrand helix, something had to hold the 2 (or 3 or 4) chains together. Their first suggestion was a "salt bridge" based on magnesium, as shown on page 62 (Norton p. 56). Remember why this was necessary. Because each of the 4 bases was a different size, putting them in the middle of the structure would (they thought) make it so irregular that it could not form the crystals that Franklin was analyzing. So they put the sugar-phosphate chains in the middle, and "solved" the problem of negative charges on the phosphates by putting a positive charged magnesium in the center. Eventually, they did get a 3-chain model that "looked good." But for the model to be truly correct, it had to be consistent with experimental data, it had to predict the X-ray data that Franklin was getting on her best pictures. In the next chapter, they met with Franklin to find out if their model might be correct, as they hoped. Note that Willy Seeds, the fellow who referred to Watson as "Honest Jim" at the beginning of the book, was among those from London who would come to Cambridge to learn about their model.
With wonderful understatement this chapter describes the nearly complete failure of the first Watson-Crick model of DNA. Their first model had no evidence to support it... and the magnesium salt bridge would not work for a molecule that was surrounded by water. Water interactions would have lifted the magnesium atom out of the bridge and broken the helix, and Franklin knew that the best X-ray patterns of DNA were given from preparations that had a high water content.
I find Watson's statement that Franklin and her student Raymond Gosling were "pugnaciously assertive" to be quite reasonable. They had just wasted an entire day traveling to Cambridge and back to view a model that had been pasted together based on unrealistic chemical assumptions and no experimental data. I would have been "assertive" too.
This chapter opens with the realization by Sir Lawrence Bragg, laboratory director, that Watson and Crick had been spending their time constructing a useless and totally unrealistic model of DNA. So, as a lab director might, he tells them to stop working it, since the London group is clearly competent enough to do the job without their "help" from Cambridge. Watson spices this momentary setback with delicious sarcasm: "After Pauling's success, no one could claim that faith in helices implied anything but an uncomplicated brain."
This chapter relates some details of Watson's personal life after Bragg's dictum to stop working on DNA, and makes it clear that he intended to do nothing of the sort. It also describes the maneuvering associated with his fellowship support from the US, and takes a backhand slap at the politics of the cold war, which he describes as "thought up by American paranoids."
At first the description of work on tobacco mosaic virus (TMV) might seem puzzling. However, TMV was the first virus to be studied, and at the time its structure was the best known of any virus. You could grow tons of it by infecting tobacco plants at your local greenhouse, then grind the leaves up and isolate more virus than you could ever hope to work on. As an added advantage, TMV could not make you sick (unless you were a tobacco plant). Therefore, it was a very convenient biological object to study. And, it had an added advantage for someone interested in DNA. It could be crystallized, and it was helical.
The "Fourier Transform" (referred to on page 76; Norton p. 69) is the mathematical operation that describes how one goes from a real object (the TMV) to its X-ray diffraction pattern, which is in mathematical terms a two-dimensional fourier transform of the object. Therefore, for any person who wishes to do serious analysis of X-ray patterns, an intimate knowledge of Fourier transforms is essential. As Watson makes clear, he hadn't a clue. Furthermore, his own attempt to produce diffraction patterns were fruitless. One again, Watson's candor reveals something about this remarkable scientist he wasn't very good at the laboratory bench.
TMV (tobacco mosaic virus)
It might seem unbelievable today to read that the greatest American chemist of his time (perhaps of all time) should have been denied a passport. I cannot think of any single fact in this book that chronicles the climate of the McCarthy era better than this one. Remember that Pauling wanted to visit England to talk about science, not politics. For those who do not know the history, Pauling was a long-time opponent of the development, testing, and deployment of nuclear weapons and protested tirelessly against them. This earned him the disapproval of the US government, but it also (eventually) earned him the Nobel Peace Prize. To this day, Linus Pauling is the only person ever to win unshared Nobel Prizes in two different fields (Chemistry and Peace).
The Hershey-Chase experiment, which is described on page 80 (Norton p. 72), was one of the key experiments that helped to establish DNA as the genetic material. In many respects it simply confirmed the earlier work of Oswald Avery on bacteria, but because the experiment was done with bacteriophage, the cool experimental system that everyone seemed interested in, it carried a great deal more weight for many scientists.
The information about DNA base composition described on page 83 (Norton pp. 75-76)is the famous "Chargaff's Rule," describing the constant ratios of bases in DNA: [A] =[ T] and [G] = [C]. If there is one factor, in my opinion, that led the Watson-Crick team to success in its quest, it was their insistence that Chargaff's rule was so important that any successful structure for DNA had to be able to explain it. As you will see, Crick had already hit upon the idea that Chargaff's rule had its basis in a base-pairing mechanism. However, the only way he could imagine to pair the bases was by stacking them directly on top of each other, something like the sketch (at right):
The nucleotide bases are very flat structures, and therefore it was possible that they might have a slight attractive force for each other based on that flatness. Crick explored this idea with theoretical chemist John Griffith. Chargaff, who visited Cambridge at this time, was not impressed with the idea, as Watson makes clear. Once again, the author's self-deprecating humor is in full view: "Only when John reassured him that I was not a typical American did he realize he was about to listen to a nut."
This chapter deals not with science but with a number of personal anecdotes and events, including the presentation of a seminar by Pauling, who finally had been able to get his passport back.
Joshua Lederberg's discovery of sexual recombination in bacteria would, just as Watson supposed, make it possible to work out the genetic organization of bacteria over the next few years. The concluding paragraph of this chapter is perhaps its most important. Here Watson describes the steady improvement in Franklin's X-ray patterns they "were getting prettier and prettier." He also notes that Franklin had begun to conclude that the sugar phosphate chains had to be on the outside of the molecule, which turned out to be an essential element of the solution to DNA's structure.
To me, one of the most interesting statements in this chapter is the claim that Watson sat down and wrote what we now call the "central dogma" on a sheet of paper, namely, that DNA makes RNA which makes protein. Towards the end of the chapter the news arrives that Linus Pauling has a new structure for DNA. Although Watson and Crick do not wish to believe it, clearly they worry that Pauling has solved the structure and the prize they hoped for is lost.
This is a key chapter. Peter Pauling, Linus' son, shows Watson and Crick a preprint of Pauling's paper describing his triple-helix model of DNA. Read the sections on page 103 (Norton p. 93) very carefully, because these describe Pauling's great mistake with his triple helix model. The key element of the model was the placement of the sugar-phosphate backbone in the center of the structure. Pauling proposed that hydrogen bonds held the strands together, and drew the structure something like this:
Unfortunately for the model, there was a problem. In order for those hydrogen bonds to form, the oxygen atoms that were part of the phosphate groups (above) would have to have hydrogen atoms bound to them. However, at any realistic cellular pH, those hydrogen atoms should have been dissociated, and the phosphate groups should have been negatively charged, like this:
This meant that Pauling's scheme was simply impossible. There is something else interesting in this passage. Watson and Crick were sure that Pauling's model was wrong (p. 104). One might think that two scientists who were sure that they had found a fatal error in another scientist's published theory would inform him of that fact. Did they? Absolutely not. Instead, they realized that they still had some time with which to win the race.
Franklin's arguments about the anti-helical features of the DNA X-ray patterns apply to a structure known as the A-form of DNA that has very little water (image at right). And it does, indeed, have anti-helical features in its diffraction pattern, features that took decades more to work out. There's a photograph of the A-form pattern among the plates in the center of the book(see figure at right). The opening pages of the chapter also carry Watson's description of a situation in which he felt that Franklin was about to strike him. Given Franklin's size and physical stature, I find this very hard to believe, and wonder about the degree to which this "incident" is a figment of Watson's imagination.
A few moments later (on page 107; Norton p. 98) a different form of DNA is described by Wilkins and shown to Watson, one known as the B-form. The B-form is clearly helical, and Franklin's patters show that. I regard page 107 as the single most important page of this remarkably book. Remember how Watson has insisted throughout that Franklin was "anti-helical" in all of their encounters and discussions. Yet, the instant that Watson, with very little understanding of crystallography, viewed the B-form, he wrote that "my mouth fell open and my pulse began to race." As he notes, the distinct black cross in the center of the structure could only arise from a helical molecule. On this page, despite his outrageous treatment of Franklin throughout the book, he reveals her true contributions to the model. She provided the critical X-ray data, she insisted that the bases be placed at the center of the structure, and she insisted that the sugar-phosphate chains were outside. The double helix model simply could not have been built without these three contributions.
The final sentence of this chapter describes another critical element of the model, namely, the decision to build models with just 2 chains. The observation that important biological objects come in pairs is just a little too cute, of course, but is right on the mark in terms of the properties of DNA.
The A-form of DNA
The B-form of DNA
Bragg, as described in the beginning of this chapter, is clearly delighted that his group will have another shot at DNA in light of Pauling's failure, and he seems to give tacit approval to Watson and Crick to go ahead. The meridional reflection at 3.4 angstroms (page 110; Norton p. 101) was a key feature of Franklin's X-ray pattern. This reflection is clearly visible on the photo of the B-form in the book, and it indicates that something repeats perpendicular to the long axis of the DNA strand at 3.4 angstroms. The simplest "somethings" that might lie in that plane are the nitrogen bases, as shown in my sketch at right:
The key was now to build a two-strand model that made chemical sense, and also explained such tidbits as Chargaff's rule.
On page 114 (Norton p. 105) note that Franklin's actual data came into W&C's hands without her knowledge. This is an important point is assessing credit (or blame) for the double helix model.
The like-with-like model that Watson describes on page 117 (Norton p. 107) is more than interesting, because (as Watson notes) a like-with-like pairing could explain how DNA replicates the two strands would each have exactly the same base sequence and he makes this explicit on page 118 (Norton p. 108). It has, however, a glaring structural flaw. Purines have two rings and are bulkier than pyrimidines (see page 42 [Norton p. 35] for drawings). If you pair like-with-like, then a purine/purine pair would be much bulkier than a pyrimidine/pyrimidine pair, making the two strands bulge alternately inward and outward. The X-ray pattern gave no hint of such irregular bulges in the molecule. Therefore the model had to be wrong.
This chapter opens with a discussion of tautomeric forms of several of the nitrogenous bases. Tautomers are chemical forms that are related to each other by the movement of a hydrogen atom. Technically, they are completely different compounds, even though they may have very similar properties. This was pointed out to Watson (who was never much of a student of chemistry) by the American crystallographer Jerry Donohue. be sure to look at the very end of the 1-page double helix paper for an acknowledgment of Donohue's help in getting the bases right.
As a result of Donohue's intervention, Watson realizes that the base pairs he had drawn for his like-with-like scheme simply wouldn't work. The model is quickly discarded.
Using the suggestion of Jerry Donohue as to the correct tautomer of each base, Watson draws out a scheme for hydrogen bonding between the bases, shown on page 124 (Norton p. 113) (Although Watson decided that the G-C base pair could only have two hydrogen bonds. It soon became clear that it actually has three).
Significantly, Francis Crick had noted a critical bit of information that helped to build their new model in a report on Franklin's work prepared for the British Medical Research Council (MRC), her funding agency, in December 1952. In that report, Franklin had noted that her data indicated that DNA had a "face-centered monoclinic unit cell." In the lingo of crystallography, that meant that the DNA molecule belonged to "space group C2."
In other words, DNA looked the same way when turned upside down. By using this information to build their model, Watson and Crick were able to determine (see page 125, Norton p. 114) that the positions that the bases must assume places them upside-down with respect to each other - implying that the two strands of the helix run in opposite directions (are antiparallel). This a key feature of the double helix model, and was directly implied by X-ray data from Franklin's lab.
Significantly, at the bottom of page 129 (Norton p. 118) Watson admits that he'd been wrong all along about the presence of Mg++ in the DNA molecule. Franklin had been right. There was no Mg++ present to act as a salt bridge. Rather, any cation present (Mg++ or Na+) would be found outside the molecule near the phosphates, and would have little effect on its structure.
The references on Page 131 (Norton p. 120) to "coordinates" are the ways in which one checks a model like this. Once one builds a model, the molecular model itself contains a number of repeat distances that should produce spots in the X-ray pattern. If you are sure that you have all of the bond lengths correct in the model, you should be able to go to the X-ray pattern and find spots in an orientation that corresponds to the repeat distances. Page 130 (Norton 119) shows one of the first schematic renderings of the double helix, emphasizing the antiparallel strands and complementary base pairing.
This chapter describes something that is often overlooked in critical accounts of the work of Watson and Crick to develop the double helix model. Once they had it, they did not conceal their success from Wilkins and Franklin, but showed it to them almost immediately. This made practical sense, since only the London lab had the actual X-ray patterns (remember, Watson didn't take or copy Franklin's data he had only been able to look at it briefly) from which measurements could be made to confirm the coordinates.
Franklin's immediate acceptance of the double helix model (p. 134, Norton p. 124) contrasts sharply with Watson's assertion that she had been "anti-helical" up to that point. And I wonder whether she was more open to the helical structure than Watson lets on. In a telling sentence, especially in light of his earlier insistence as to how "stubborn" Franklin had been, on page 134 (Norton p. 124) Watson admits that Franklin had been right all along in placing the sugar-phosphate chains on the outside of the model. Her assertion, he writes, was "foolproof," and thus reflected "first rate science."
This chapter describes some of the details of the publication of reports of the data and the model in the journal NATURE. The Watson-Crick paper, which was just a single page in length, concluded with what many critics have regarded as one of the most remarkable sentences ever put into a scientific paper: "It has not escaped out notice that the specific base pairing we have postulated immediately suggests a copying mechanism." It does indeed, and this was the beauty of the model, and one of the principal reasons why it was so quickly accepted.
There is a striking contrast between this epilogue and the treatment of Franklin in the book. Watson goes so far as to say that much of what he had written "in the early pages of this book" was wrong. A remarkable statement. While I do not know the inside story of this epilogue, it is worth noting that this manuscript was rejected by Harvard University Press because of its blatant unfairness to Franklin. When he was finally able to publish it, I've been told that his editors insisted upon the epilogue to help mute the obvious unfairness to Franklin.
Copyright 2009, Kenneth R. Miller