Proteins are thus a family of subtle and versatile molecules. As soon as I learned about them I realized that one of the key problems was to explain how they were synthesized.
There was a third important generalization, though in the 1940s this was sufficiently new that not everybody was inclined to accept it. This idea was due to George Beadle and Ed Tatum. (They too were to receive a Nobel Prize, in 1958, for their discovery.) Working with the little bread-mold Neurospora, they had found that each mutant of it they studied appeared to lack just a single enzyme. They coined the famous slogan “One gene—one enzyme.”
Thus the general plan of living things seemed almost obvious. Each gene determines a particular protein. Some of these proteins are used to form structures or to carry signals, while many of them are the catalysts that decide what chemical reactions should and should not take place in each cell. Almost every cell in our bodies has a complete set of genes within it, and this chemical program directs how each cell metabolizes, grows, and interacts with its neighbors. Armed with all this (to me) new knowledge, it did not take much to recognize the key questions. What are genes made of? How are they copied exactly? And how do they control, or at least influence, the synthesis of proteins?
It had been known for some time that most of a cell’s genes are located on its chromosomes and that chromosomes were probably made of nucleoprotein—that is, of protein and DNA, with perhaps some RNA as well. In the early 1940s it was thought, quite erroneously, that DNA molecules were small and, even more erroneously, simple. Phoebus Levene, the leading expert on nucleic acid in the 1930s, had proposed that they had a regular repeating structure [the so-called tetranucleotide hypothesis]. This hardly suggested that they could easily carry genetic information. Surely, it was thought, if genes had to have such remarkable properties, they must be made of proteins, since proteins as a class were known to be capable of such remarkable functions. Perhaps the DNA there had some associated function, such as acting as a scaffold for the more sophisticated proteins.
It was also known that each protein was a polymer. That is, it consisted of a long chain, known as a polypeptide chain, constructed by stringing together, end to end, small organic molecules, called monomers since they are the elements of a polymer. In a homopolymer, such as nylon, the small monomers are usually all the same. Proteins are not as simple as that. Each protein is a heteropolymer, its chains being strung together from a selection of somewhat different small molecules, in this instance amino acids. The net result is that, chemically speaking, each polypeptide chain has a completely regular backbone, with little side-chains attached at regular intervals. It was believed that there were about twenty different possible side-chains (the exact number was not known at that time). The amino acids (the monomers) are just like the letters in a font of type. The base of each kind of letter from the font is always the same, so that it can fit into the grooves that hold the assembled type, but the top of each letter is different, so that a particular letter will be printed from it. Each protein has a characteristic number of amino acids, usually several hundred of them, so any particular protein could be thought of crudely as a paragraph written in a special language having about twenty (chemical) letters. It was not then known for certain, as it is now, that for each protein the letters have to be in a particular order (as indeed they have to be in a particular paragraph). This was first shown a little later by the biochemist Fred Sanger, but it was easy enough to guess that this was likely to be true.
Of course each paragraph in our language is really one long line of letters. For convenience this is split up into a series of lines, written one under the other, but this is only a secondary matter, since the meaning is exactly the same whether the lines are long or short, few or many, provided we take care about splitting the words at the end of each line. Proteins were known to be very different. Although the polypeptide backbone is chemically regular, it contains flexible links, so that in principle many different three-dimensional shapes are possible. Nevertheless, each protein appeared to have its own shape, and in many cases this shape was known to be fairly compact (the word used was “globular”) rather than very extended (or “fibrous”). A number of proteins had been crystallized, and these crystals gave detailed X-ray diffraction patterns, suggesting that the three-dimensional structure of each molecule of a particular kind of protein was exactly (or almost exactly) the same. Moreover many proteins, if heated briefly to the boiling point of water, or even to some temperature below this, became denatured, as if they had unfolded so that their three-dimensional structure had been partly destroyed. When this happened the denatured protein usually lost its catalytic or other function, strongly suggesting that the function of such a protein depended on its exact three-dimensional structure.
And now we can approach the baffling problem that appeared to face us. If genes are made of protein, it seemed likely that each gene had to have a special three-dimensional, somewhat compact structure. Now, a vital property of a gene was that it could be copied exactly for generation after generation, with only occasional mistakes. What we were trying to guess was the general nature of this copying mechanism. Surely the way to copy something was to make a complementary structure—a mold—and then to make a further complementary structure of the mold, to produce in this way an exact copy of the original. This, after all, is how, broadly speaking, sculpture is copied. But then the dilemma arose: It is easy to copy the outside of a three-dimensional structure in this way, but how on earth could one copy the inside? The whole process seemed so utterly mysterious that one hardly knew how to begin thinking about it.
Of course, now that we know the answer, it all seems so completely obvious that no one nowadays remembers just how puzzling the problem seemed then. If by chance you do not know the answer, I ask you to pause a moment and reflect on what the answer might be. There is no need, at this stage, to bother about the details of the chemistry. It is the principle of the idea that matters. The problem was not made easier by the fact that many of the properties of proteins and genes just outlined were not known for certain. All of them were plausible and most of them seemed very probable but, as in most problems near the frontiers of research, there were always nagging doubts that one or more of these assumptions might be dangerously misleading. In research the front line is almost always in a fog.
So what was the answer? Curiously enough, I had arrived at the correct solution before Jim Watson and I discovered the double-helical structure of DNA. The basic idea (which was not entirely new) was this: All a gene had to do was to get the sequence of the amino acids correct in that protein. Once the correct polypeptide chain had been synthesized, with all its side chains in the right order, then, following the laws of chemistry, the protein would fold itself up correctly into a unique three-dimensional structure. (What the exact three-dimensional structure of each protein was remained to be determined.) By this bold assumption the problem was changed from a three-dimensional one to a one-dimensional one, and the original dilemma largely disappeared.
Of course, this had not solved the problem. It had merely transformed it from an intractable one to a manageable one. For the problem still remained: how to make an exact copy of a one-dimensional sequence. To approach that we must return to what was known about DNA.
By the late 1940s our knowledge of DNA had improved in several important respects. It had been discovered that DNA molecules were not, after all, very short. Exactly how long they were was not clear. We know now that they appeared to be short because, being long molecules (in the sense that a piece of string is long), they could easily be broken in the process of getting them out of the cell and manipulating them in the test tube. Just stirring a DNA solution is enough to break the longer molecules. Their chemistry was now known more correctly, and moreover the tetranucleotide hypothesis was dead, killed by some very beautiful work by a chemist at Columbia, the Austrian refugee Erwin Chargaff. DNA was known to be a polymer, but with a very different backbone and with onl
y four letters in its alphabet, rather than twenty. Chargaff showed that DNA from different sources had rather different amounts of those four bases (as they were called). Perhaps DNA was not such a dumb molecule after all. It might conceivably be long enough and varied enough to carry some genetic information.
Even before I left the Admiralty there had been some quite unexpected evidence pointing to DNA as near the center of the mystery. In 1944 Avery, MacLeod, and McCarty, who worked at the Rockefeller Institute in New York, had published a paper claiming that the “transforming factor” of pneumococcus consisted of pure DNA. The transforming factor was a chemical extracted from a strain of bacteria having a smooth coat. When added to a related strain lacking such a coat it “transformed” it, so that some of the recipient bacteria acquired the smooth coat. More important, all the descendants of such cells had the same smooth coat. In the paper the authors were rather cautious in interpreting their result, but in a now-famous letter to his brother Avery expressed himself more freely. “Sounds like a virus—may be a gene,” he wrote.
This conclusion was not immediately accepted. An influential biochemist, Alfred Mirsky, also at the Rockefeller, was convinced that it was an impurity of the DNA that was causing the transformation. Subsequently more careful work by Rollin Hotchkiss at the Rockefeller showed that this was highly unlikely. It was argued that Avery, MacLeod, and McCarty’s evidence was flimsy, in that only one character had been transformed. Hotchkiss showed that another character could also be transformed. The fact that these transformations were often unreliable, tricky to perform, and only altered a minority of cells did not help matters. Another objection was that the process had been shown to occur just in these particular bacteria. Moreover, at that time no bacterium of any sort had been shown to have genes, though this was discovered not long afterward by Joshua Lederberg and Ed Tatum. In short, it was feared that transformation might be a freak case and misleading as far as higher organisms were concerned. This was not a wholly unreasonable point of view. A single isolated bit of evidence, however striking, is always open to doubt. It is the accumulation of several different lines of evidence that is compelling.
It is sometimes claimed that the work of Avery and his colleagues was ignored and neglected. Naturally there was a mixed spectrum of reactions to their results, but one can hardly say no one knew about it. For example, that august and somewhat conservative body, the Royal Society of London, awarded the Copley Medal to Avery in 1945, specifically citing his work on the transforming factor. I would dearly love to know who wrote the citation for them.
Nevertheless, even if all the objections and reservations are brushed aside, the fact that the transforming factor was pure DNA does not in itself prove that DNA alone is the genetic material in pneumococcus. One could quite logically claim that a gene there was made of DNA and protein, each carrying part of the genetic information, and it was just an accident of the system that in transformation the altered DNA part was carrying the information to change the polysaccharide coat. Perhaps in another experiment a protein component might be found that would also produce a heritable change in the coat or in other cell properties.
Whatever the interpretation, because of this experiment and because of the increased knowledge of the chemistry of DNA, it was now possible that genes might be made of DNA alone. Meanwhile the main interest of the group at the Cavendish was in the three-dimensional structure of proteins such as hemoglobin and myoglobin.
4
Rocking the Boat
LET US NOW return to my own career. I still had to make contact with Max Perutz. One day in the late 1940s, I was returning to Cambridge from a visit to London, having arranged to call on Perutz at the physics laboratory where he worked. The train journey from London was uneventful. I watched the countryside slide past but my thoughts were elsewhere, focused mainly on my impending visit to the Cavendish Laboratory. For a British physicist the Cavendish had a unique glamour. It had been named after the eighteenth-century physicist Henry Cavendish, a recluse and an experimenter of genius. The first professor had been the Scottish theoretical physicist James Clerk Maxwell, of Maxwell’s equations. While the laboratory was being built he did experiments in his kitchen at home, his wife raising the room temperature for him by boiling pans of water.
It was at the Cavendish that J. J. Thomson had “discovered” the electron by making measurements of both its mass and its charge. Thompson was an interesting case of an experimenter who was so clumsy that his associates tried to keep him away from his own apparatus, for fear of his breaking it. Ernest Rutherford, fresh from New Zealand, had started his main research career there and later returned to succeed J.J. as Cavendish Professor. There, under his direction Cockroft and Walton had first “smashed the atom”—that is, had produced the first artificial atomic disintegration. Their original accelerator was still there. And in the early 1930s James Chadwick (whom I knew later as Master of Caius College) had in a few short weeks discovered the neutron. At that time the Cavendish was in the very forefront of research in fundamental physics.
The current Cavendish Professor was Sir Lawrence Bragg (known to his close friends as Willie), the formulator of Bragg’s law for X-ray diffraction. He was the youngest Nobel Prize winner ever, having been only twenty-five when he shared it with his father, Sir William Bragg. It was no wonder that I was in awe of such a world-famous institution and excited at the prospect of visiting it.
At the station I decided to take a taxi. After settling my bags, I leaned back in my seat. “Take me,” I said, “to the Cavendish Laboratory.”
The driver turned his head to look at me over his shoulder. “Where’s that?” he asked.
I realized, not for the first time, that not everyone was as deeply interested in fundamental science as I was. After fumbling in my papers I found the address.
“It’s in Free School Lane,” I said, “wherever that is.”
“Not far from the Market Square,” said the cabby, and off we went.
Max Perutz, whom I was to visit, was Austrian by birth. He had obtained his first degree, in chemistry, at the University of Vienna. He had wanted to go to Cambridge to work under Gowland Hopkins, the founder of the Cambridge School of Biochemistry. Perutz had asked Herman Mark, the polymer specialist, to try to arrange this for him when Mark went on a short visit to Cambridge. Instead Mark ran into J. D. Bernal (known to his close friends as “Sage,” because he appeared to know everything). Bernal said he would be happy to have Perutz work with him and so Max became a crystallographer. This was all before the Second World War.
By the time of my visit Perutz was working, under the loose supervision of Bragg, on the three-dimensional structures of proteins. As I explained in the last chapter, proteins belong to one of the key families of biological macromolecules. How each protein acts depends on its exact three-dimensional structure. It is therefore crucially important to discover such structures experimentally. At that time the largest organic molecule whose three-dimensional structure had been determined by X-ray diffraction was two orders of magnitude smaller than a typical protein. A determination of the three-dimensional structure of a protein seemed, to most crystallographers, almost impossible or, at best, very far away. Bernal had always been enthusiastic about it, but then he was a visionary. However, it also had a great appeal for the hard-headed Bragg, since it represented a challenge. Having started his career unraveling the very simple structure of crystals of sodium chloride (common table salt), Bragg hoped he might crown his achievements by solving one of the largest possible molecular structures.
Before the war, Bernal had founded the study of the X-ray diffraction of protein crystals. One day, he was observing the optical properties of a protein crystal, using the light microscope (actually a polarizing microscope). The crystal was sitting on an open glass slide, with a little bit of the mother liquor of the crystal (the solution in which the protein crystal had been grown) attached to it. Slowly the water in the mother liquor evaporated into
the air till eventually the crystal became dry. As it did so Bernal saw the optical properties deteriorate, since the dry jumbled crystal transmitted the light in a more confused way than before. Bernal immediately realized that it was important to keep protein crystals wet and proceeded to mount a crystal in a small silica tube, sealed with a special wax at each end. Fortunately the silica interfered very little with the X rays being diffracted from the crystal. All previous attempts to get X-ray diffraction photos from protein crystals had produced only a few smudges on the photographic plate since the crystals used had dried in the air. Great was the excitement in Bernal’s lab when the wet crystal produced many beautiful spots. The study of protein structure had taken a decisive first step.
Before I first visited Max Perutz at the Cavendish I read the two papers he had recently published in the Proceedings of the Royal Society about his X-ray diffraction studies on crystals of a variety of hemoglobin. Hemoglobin is the protein that carries oxygen in our blood and makes red blood cells red, though the variety Perutz had studied came from a horse, as horse hemoglobin happens to form crystals that are especially convenient for X-ray studies. We now know that each hemoglobin molecule is made up of four rather similar subunits, each of which contains about 2, 500 atoms, arranged in a precise three-dimensional structure.
Since one cannot easily focus X rays, it is impossible to make X-ray photographs in the way one uses a lens to make photographs using visible light or by focusing electrons in the electron microscope. However, the wavelength of convenient X rays is about the same distance as the distance between close atoms in an organic molecule. For this reason the pattern of X rays that molecules scatter can, under optimal circumstances, contain enough information for the experimenter to determine the positions of all the atoms in the molecule. More correctly, such a picture shows the density of the electrons that surround each atom and that, since they have very little mass, scatter the X rays more effectively than the heavier atomic nuclei. A crystal is used because the X rays scattered from a single molecule would be too feeble. If long exposures were used to try to overcome this difficulty, the heavy dose of X rays would damage the molecule far too much and effectively cook it before enough X rays had been scattered to be useful.
What Mad Pursuit Page 5