CHAPTER FOUR—VICTORY ON PAPER

  The key to the answer to the problem of protein structure was found by a Russian. This was Michael Tswett.

  In 1906, Tswett submitted a paper to a German botanical journal in which he described a series of experiments involving a new, and, as it turned out, revolutionary technique. Tswett was a botanist who was interested in the Colored pigments one could soak out of plant leaves by using various solvents. Among those pigments is chlorophyll which plants use to convert solar energy into food and without which life on Earth—except for certain micro-organisms—would quickly become impossible. Naturally, biochemists were yearning at the time to get at those plant pigments, separate one from another and figure out the structure of each. But how was one to go about separating the unholy mess into individual components? Ordinary chemical procedures simply didn’t come close to doing the job.

  The way Tswett went about it was to dissolve this mess in a liquid called petroleum ether and then pour it through a glass column packed tightly with powdered limestone. The liquid percolated downward and came out at the bottom of the column unchanged and unharmed. The plant pigments which had been dissolved in the liquid, however, remained behind, clinging to the surface of the limestone particles.

  Don’t think for a moment that the pigment molecules were faced with an easy choice. To be sure they preferred the particle surface, yet the liquid did exert a certain attraction for them. As the liquid passed through the pigment, molecules were slowly and reluctantly dragged down with it, moving downward from limestone particle to limestone particle.

  Each individual type of pigment was its own schizophrenic self; each arrived at its own particular compromise in deciding how firmly to remain with the limestone or how willingly accompany the downward-moving liquid. The more tightly they hugged the particles, the more slowly that particular variety of molecule moved downward. The more bibulously they enjoyed their liquid surrounding, the more quickly they moved down.

  What was, therefore, originally a disheartening mixture slowly resolved itself into a series of bands of different shades of yellow and green in different places along the column of powdered limestone (see Figure 10). If one continued to pour petroleum ether through the column, each band would eventually he washed out through the bottom opening—one at a time. By the time the experiment was completed, the different components of the mixture would be resting contentedly in separate beakers.

  Tswett called the technique ‘chromatography’ from Greek words meaning ‘color-writing’, though as he pointed out, the principle would work for colorless mixtures as well.

  Tswett, unfortunately, was in a poor position. Biochemistry was, at the time, almost the private domain of German scientists and these did not take kindly to the fact that here was a neat, elegant and easy solution to a mystifying and tantalizing problem offered the world by (a) a botanist and not a biochemist, and (b) a Russian and not a German. Furthermore, in 1910, when Tswett wrote a detailed monograph on chromatography, he wrote in the very best Russian and he might as well have used south-Martian for all the good that did the biochemical world.

  On the side of Tswett was only the fact that he was right and that chromatography was destined to become one of the most powerful and widely-used techniques available to the biochemist. The mere fact of his being right, however, was not enough to raise a Russian botanist to a level of equality with a German biochemist, and chromatography dropped dead. (In 1922, an American used chromatography and reported it, but in those days that carried little weight, too.)

  Twenty-five years passed from the day of the original discovery. Then, in 1931, German biochemists finally got around to using Tswett’s techniques, and what do you know, it worked exactly as he had described.

  In the last quarter-century, all sorts of powders have been used to separate individual components out of all sorts of mixtures. Most recently, synthetic substances known as ‘ion-exchange resins’ have been most useful.

  In 1944, came a major refinement. A group of English biochemists abandoned columns and powders and contented themselves with sheets of filter paper (i.e. a kind of porous paper which, in its better grades, is almost pure cellulose).

  If one end of a strip or sheet of filter paper is immersed in liquid, the liquid will slowly creep up the filter paper. (You can watch this phenomenon yourself if you have a piece of blotting paper and a bottle of ink handy.) If you keep the filter paper and the liquid in a closed container to prevent evaporation, the liquid will eventually soak through the entire strip if it is not too long.

  Now suppose that near the end of a sheet of filter paper you were to place a drop or two of a solution containing a mixture of similar substances and then let the drop dry. Next, dip that end of the sheet into a liquid, being careful to keep the dried drop of mixture above the level of the liquid, and enclose the whole system to cut down evaporation.

  Up creeps the liquid. In a short while it reaches and passes the dried drop of mixed substances. Each different component of that mixture is now faced with the usual schizophrenic dilemma. Shall it stay put or shall it let go? Shall it ignore the liquid or shall it go along with it? Each substance makes the usual individual compromise. Each substance moves along with the liquid in a laggard and hesitant way. (Referring back to a home experiment with a blotter and a drop of ink, note that the pigment particles in the ink do not travel as far along the blotting paper as does the water content of the ink, so that the blue drop of spread-out ink is encircled by a colorless damp spot.)

  As you can guess, each component of the mixture travels at its own rate. By the time the liquid has soaked a foot or two along the paper, the original spot has become a whole series of spots.

  Theoretically, each spot of the series should now contain a single separate substance. Actually, however, in any mixture containing a number of similar substances, it often happens that two or three may have such similar rates of travel that at the end, they remain in a single spot.

  For that reason, the paper is dried, turned on its side, and immersed in a different type of liquid altogether. The first liquid, for instance, might have been a mixture of butyl alcohol and water; the second, a mixture of phenol and water. Substances which have similar rates of travel in one liquid are very likely to have different rates of travel in a second liquid. The two or three substances, which had previously stuck together buddy-fashion, bid one another a fond adieu and separate. The whole process is diagrammed (Figure 11).

  This technique is called two-dimensional paper chromatography. Its advantage over the earlier column chromatography is that the equipment needed is dirt-cheap and that very small quantities of material can be handled without difficulty.

  Once the spots are separated; once each substance occupies its own individual place on the paper; there is the problem of finding them. Usually, the substances being separated are colorless and the filter paper, after drying, has a disconcertingly virginal emptiness about it.

  The problem can be solved handily, however. For instance, under ultraviolet light, the spots may fluoresce or appear black. In either case, pencil lines can be drawn about them. Or else, two sheets are prepared in identical fashion and one is treated with some chemical which will combine with some or all of the substances to form a visible color. The spots stand out neatly; the colored sheet is superimposed on the untouched sheet, and you take it from there.

  Once the spots are located, they can be cut out. Each substance can be dissolved individually out of the paper. Each can be identified and manipulated further. Peace, it’s wonderful.

  And why am I talking about this? Where are all these filter paper manipulations getting me?

  Well, it is paper chromatography which enabled chemists to solve the problem which in the previous chapter I went to great lengths to demonstrate to be ‘impossible’ of solution. That problem involves the structure of protein molecules.

  Each protein molecule is made up of hundreds or even thousands of simpler substances called amino-acids—of some twenty different varieties—which are strung together like pearls in a necklace. The number of different ways in which a particular combination of amino-acids can be arranged even for a protein of only average size is so great that all of space and time—literally, not poetically, speaking—is insufficient to allow each possible way to be tested in order that the one arrangement which actually makes up that protein be discovered.

  Nor is this ‘impossible’ problem just a matter of idle curiosity on the part of long-haired biochemists who have nothing better to do.

  In case you wonder about that, let’s bring insulin to the front of the stage.

  Insulin is a protein molecule which is manufactured by certain specialized cells of the pancreas (a gland located just under the stomach). As it is formed, it is secreted into the blood in amounts adjusted to the needs of the body at the moment. The blood carries it to all the cells of the body and there it somehow supervises the utilization of sugars and fats for energy-production purposes.

  Exactly how insulin does this is a matter of considerable dispute. Some biochemists think it acts as a control on one particular chemical reaction, through which the entire series of reactions is hastened or slowed according to need. Other biochemists think insulin coats the surface of each cell and controls the flow of raw materials entering, adjusting the cell chemistry in that fashion.

  Whatever the exact mechanism, insulin is vital. Every once in a while, the pancreas stops manufacturing this key protein in some individual. The chemistry of the body promptly goes wrong. Glucose—a kind of sugar used by the body for quick energy production—is processed inefficiently. It accumulates in the blood and spills over into the urine. Sugar in the urine or, better still, too much sugar in the blood, is an almost certain sign of the disease called diabetes.

  Because a diabetic utilizes his food inefficiently, he grows hungrier; yet though he may increase his food intake, he will lose weight nevertheless. He needs extra water to carry off the sugar continually passing through his kidneys, so he must drink more and urinate more. The disease has its ramifications. The diabetic is more prone to various infections than is the normal person, he is much more likely to suffer from hardening of the arteries if the disease is allowed to take its course.

  Although diabetes tends to run in families, its onset in an individual is unpredictable and unpreventable. Once it comes, it is incurable. (Careful diet may delay its approach and keep its effects relatively mild.) Diabetes is the most common chemical disorder of the human body. Millions suffer from this serious disease.

  Fortunately, in the 1920s, some Canadian scientists—who later got the Nobel Prize for it—discovered how to isolate insulin from the pancreases of cattle. Using such insulin as a replacement for that which their own pancreases can no longer supply, human diabetics can now live reasonably normal lives.

  The use of insulin as a treatment (not cure) for diabetes has certain difficulties about it. First, its only source is the pancreas of slaughtered cattle, swine and so on, and each animal has but one pancreas. There is, therefore, an upper limit to how much insulin can be made available. Secondly, insulin cannot be taken by mouth, since it is digested and made useless in the stomach and intestines. It must be injected by hypodermic needle, which is troublesome.

  Now if the exact structure of insulin were known—not just the approximate structure but the exact structure—biochemists might be better able to decide from that structure its method of working, now under such dispute. They might make an intelligent guess at what features of its molecule were most necessary for its purpose and synthesize a simpler molecule containing those features. If the simpler molecules worked to control diabetes, it would mean that there would be a potentially limitless supply of drug not dependent on cattle. Furthermore, it might be simple enough to withstand digestion, in which case it might be taken by mouth.

  This sort of procedure has actually been carried out in the case of certain alkaloids. The structure of the local anesthetic cocaine was worked out and simpler substances, containing the essential features of the molecule, were synthesized. Such a synthetic substitute-cocaine is Novocaine which, in some respects, is more useful than the natural drug.

  So you see then that in the case of insulin, at least, the exact arrangement of the amino-acids is anything but an academic problem. It has an important application to a serious health problem.

  The size of the insulin molecule can be determined in a number of ways and it is found to have a molecular weight of 12,000. This is 660 times as great as the weight of a water molecule but only one-fifth the weight of an averagely-sized protein such as hemoglobin. Despite its small size for a protein, insulin still has room for about a hundred amino-acid components, which makes the problem of its exact structure a sizable one.

  The insulin molecule can be broken up into the individual amino-acid components by prolonged treatment with acid. Before 1944 this wouldn’t have helped much because many of the amino-acids are quite similar in structure and it is the devil’s own job to tackle the analysis of amino-acid mixtures in the expectation of determining how much of each amino-acid is present. With paper chromatography, however, the problem is simple. A drop of the mixture is placed on the filter paper, two different solvents are used in two different directions, and the various amino-acids are spread out neatly so that the identity of each and the quantity present can be determined.

  In this way, it was found that the molecule of insulin contained 96 amino-acids of 18 different types. For our purposes, the names of the different amino-acids are unnecessary. We can list them in alphabetical order and call them A, B, C, through R.

  The fifth amino-acid—in alphabetical order—is different from the rest in that it is a double molecule, or a two-headed molecule if you prefer. One end of it can form part of one amino-acid chain and the other end of it can form part of a second amino-acid chain. For that reason, it will be referred to as E-E, instead of simply E.

  Table XIII lists the different amino-acids and gives the number of each which is found in the insulin molecule. The number of ways in which those ninety-six amino-acids can be arranged in a chain to form a protein molecule is three googols; that is, 3 x 10100 or a 3 followed by 100 zeros. I won’t go through the gyrations I went through in a previous chapter to prove that this is a large number. Take my word for it. The total number of all subatomic particles contained in a trillion suns is nothing in comparison to it.

  Which of the three googols of possible arrangements is the right one? Give up? Well, a group of British biochemists under the direction of Dr. F. Sanger didn’t. They began working on the problem in 1945 or thereabouts.

  One point of attack in any amino-acid chain is the end amino-acids. Suppose you had the chain, F-G-H-I-J-K. Obviously, F and K differ from the other amino-acids in that each has one end free. F has its acid side free and K has its amino side free. (Arbitrarily, I write the chain so that the acid ends of each amino-acid component is at the left and the amino ends of each at the right. It could be done the other way around just as easily.)

  Sanger and his groups discovered that if an amino-acid chain is treated with a certain colored chemical—which is now called Sanger’s Reagent after him—it will attach itself to the unattached amino group at the extreme right-hand end of the chain. You would have this situation in the case we have presented: F-G-H-I-J-KS, where S represents Sanger’s Reagent.

  If, after treatment, the chain is broken into individual amino-acids by acid treatment, Sanger’s Reagent remains combined and you’re left with F, G, H, I, J, and KS. The mixture can be chromatographed and the KS is extremely easy to locate since, like Sanger’s Reagent alone, it is colored and the other amino-acids are not. The KS can be dissolved out of the paper, Sanger’s Reagent can be forced off and the amino-acid identified. In this way, one can decide the particular amino-acid which exists at the extreme right end of an amino-acid chain.

  Sanger’s group applied this principle to insulin and found that every molecule of insulin yielded four amino-acids to which Sanger’s Reagent was attached. Two of these amino-acids were H and two were M (using our alphabetical arrangement).

  The only conclusion was that every insulin molecule consists of four separate amino-acid chains held together by the double-headed amino-acid E-E of which there are six in every molecule. The picture so far is shown diagrammatically in Figure 12.

  Now there is a way of breaking the double-headed E-E into two single-headed fragments, E and E, without disturbing other portions of the amino-acid chains. The chemical used is one called performic acid.

  Sanger and company treated insulin with performic acid—what an opportunity for puns—and found themselves left with the four isolated amino-acid chains shown in Figure 13.

  The two chains ending in H and containing four E’s apiece turned out to be identical, judging from the results of various tests. Let’s refer to such a chain as Chain I. The two chains ending in M and containing two E’s apiece are also identical. Call such a chain Chain II. Since Chain I and Chain II are different in structure, they have different chemical properties and can be separated easily enough.

  Once separated, Chain I and Chain II can be separately broken up into individual amino-acids by treatment with acid. The resulting amino-acid soup from each type of chain can be and was analyzed on paper chromatography. In this way, the different amino-acids in each chain can be identified both as to nature and quantity. The results are shown in Table XIV.

  Notice that Chain I consists of twenty-one amino-acids and Chain II of thirty amino-acids. Since each insulin molecule consists of two of Chain I and two of Chain II, the total number of amino-acids in insulin comes to one hundred and two. Earlier, I said ninety-six. This is no discrepancy, however, since in breaking apart the four amino-acid chains of insulin, the six E-E amino-acids were converted to twelve E amino-acids, thus adding six amino-acids to the total, 96 plus 6 equals 102, Q. E. D.