by Isaac Asimov
In this shorthand, the formula of serum albumin can be written: Gly15Val45Leu58Ileu9Pro31Phe33Tyr18Try1Ser22Thr27CyS32CySH4Met6Arg25His16Lys58Asp46Glu80—more concise, you will admit, though certainly nothing to be rattled off.
ANALYZING THE PEPTIDE CHAIN
Discovering the empirical formula of a protein was only half the battle—in fact, much less than half. Now came the far more difficult task of deciphering the structure of a protein molecule. There was every reason to believe that the properties of every protein depend on exactly how—in what order—all those amino acids are arranged in the molecular chain. This assumption presents the biochemist with a staggering problem. The number of possible arrangements in which nineteen amino acids can be placed in a chain (even assuming that only one of each is used) comes to nearly 120 million billion. If you find this hard to believe, try multiplying out 19 times 18 times 17 times 16, and so on—the way the number of possible arrangements is calculated. And if you do not trust the arithmetic, get nineteen checkers, number them 1 to 19, and see in how many different orders you can arrange them. I guarantee you will not continue the game long.
When you have a protein of the size of serum albumin, composed of more than 500 amino acids, the number of possible arrangements comes out to something like 10600—that is, 1 followed by 600 zeros. This is a completely fantastic number—far more than the number of subatomic particles in the entire known universe—or, for that matter, far more than the universe could hold if it were packed solid with such particles.
Nevertheless; although it may seem hopeless to find out which one of all those possible arrangements a serum albumin molecule actually possesses, this sort of problem has actually been tackled and solved.
In 1945, the British biochemist Frederick Sanger set out to determine the order of amino acids in a peptide chain. He started by trying to identify the amino acid at one end of the chain—the amine end.
Obviously, the amine group of this end amino acid (called the N-terminal amino acid) is free—that is, not attached to another amino acid. Sanger made use of a chemical that combines with a free amine group but not with an amine group bound to a carboxyl group and produces a DNP (dinitrophenyl) derivative of the peptide chain. With DNP he could label the N-terminal amino acid, and since the bond holding this combination together is stronger than the bonds linking the amino acids in the chain, he could break up the chain into its individual amino acids and isolate the one with the DNP label. As it happens, the DNP group has a yellow color, so this particular amino acid, with its DNP label, shows up as a yellow spot on a paper chromatogram.
Thus, Sanger was able to separate and identify the amino acid at the amine end of a peptide chain. In a similar way, he identified the amino acid at the other end of the chain—the one with a free carboxyl group, called the C-terminal amino acid. He was also able to peel off a few other amino acids one by one and identify the end sequence of a peptide chain in several cases.
Now Sanger proceeded to attack the peptide chain all along its length. He worked with insulin, a protein that has the merit of being very important to the functioning of the body and the added virtue of being rather small for a protein, having a molecular weight of only 6,000 in its simplest form. DNP treatment showed this molecule to consist of two peptide chains, for it contains two different N-terminal amino acids. The two chains are joined by cystine molecules. By a chemical treatment that broke the bond between the two sulfur atoms in the cystine, Sanger split the insulin molecule into its two peptide chains, each intact. One of the chains had glycine as the N-terminal amino acid (call it the G-chain), and the other had phenylalanine as the N-terminal amino acid (the P-chain). The two could now be worked on separately.
Sanger and a co-worker, Hans Tuppy, first broke up the chains into individual amino acids and identified the twenty-one amino acids that make up the G-chain and the thirty that compose the P-chain. Next, to learn some of the sequences, they broke the chains, not into individual amino acids, but into fragments consisting of two or three. This task could be done by partial hydrolysis, breaking only the weaker bonds in the chain, or by attacking the insulin with certain digestive substances which broke only certain links between amino acids and left the others intact.
By these devices Sanger and Tuppy broke each of the chains into many different pieces. For instance, the P-chain yielded 48 different fragments, 22 of which were made up of two amino acids (dipeptides), 14 of three, and 12 of more than three.
The various small peptides, after being separated, could then be broken down into their individual amino acids by paper. chromatography. Now the investigators were ready to determine the order of the amino acids in these fragments. Suppose they had a dipeptide consisting of valine and isoleucine. The question would be: Was the order Val-lieu or Ileu-Val? In other words, was valine or isoleucine the N-terminal amino acid? (The amine group, and consequently the N-terminal unit, is conventionally considered to be at the left end of a chain.) Here the DNP label could provide the answer. If it was present on the valine, that would be the N-terminal amino acid, and the arrangement in the dipeptide would then be established to be Val-lieu. If it was present on the isoleucine, it would be Ileu-Val,
The arrangement in a fragment consisting of three amino acids also could be worked out. Say its components were leucine, valine, and glutamic acid. The DNP test could first identify the N-terminal amino acid. If it was, say, leucine, the order had to be either Leu-Val-Clu or Leu-Clu-Val. Each of these combinations was then synthesized and deposited as a spot on a chromatogram to see which would occupy the same place on the paper as did the fragment being studied.
As for peptides of more than three amino acids, these could be broken down to smaller fragments for analysis.
After thus determining the structures of all the fragments into which the insulin molecule had been divided, the next step was to put the pieces together in the right order in the chain-in the fashion of a jigsaw puzzle. There were a number of clues. For instance, the G-chain was known to contain only one unit of the amino acid alanine. In the mixture of peptides obtained from the breakdown of G-chains, alanine was found in two combinations: alanine-serine and cystine-alanine. Hence, in the intact G-chain, the order must be CyS-Ala-Ser.
By means of such clues, Sanger and Tuppy gradually put the pieces together. It took a couple of years to identify all the fragments definitely and arrange them in a completely satisfactory sequence; but by 1952, they had worked out the exact arrangement of all the amino acids in the G-chain and the P-chain. They then went on to establish how the two chains were joined. In 1953, their final triumph in deciphering the structure of insulin was announced. The complete structure of an important protein molecule had been worked out for the first time. For this achievement, Sanger was awarded the Nobel Prize in chemistry in 1958.
Biochemists immediately adopted Sanger’s methods to determine the structure of other protein molecules. Ribonuclease, a protein molecule consisting of a single peptide chain with 124 amino acids, was conquered in 1959; and the protein unit of tobacco mosaic virus, with 158 amino acids, in 1960. In 1964, trypsin, a protein with 223 amino acids, was deciphered. By 1967, the technique was actually automated. The Swedish-Australian biochemist Pehr Edman devised a sequenator which could work on 5 milligrams of pure protein, peeling off and identifying the amino acids one by one. Sixty amino acids of the myoglobin chain were identified in this fashion in four days.
Ever longer peptide chains have been worked out in full detail; and by the 1980s, it was quite certain that the detailed structure of any protein, however large, could be determined. It was only necessary to take the trouble.
In general, such analyses have shown that most proteins have all the various amino acids (or almost all) well represented along the chain. Only a few of the simpler fibrous proteins, such as those found in silk or in tendons, are heavily weighted with two or three amino acids.
In those proteins ma
de up of all nineteen amino acids, the individual amino acids are lined up in no obvious order; there are no easily spotted periodic repetitions. Instead, the amino acids are so arranged that when the chain folds up through the formation of hydrogen bonds here and there, various side chains make up a surface containing the proper arrangement of atomic groupings or of electric-charge pattern to enable the protein to do its work.
SYNTHETIC PROTEINS
Once the amino-acid order in a polypeptide chain was worked out, it became possible to attempt to put together amino acids in just that right order. Naturally, the beginning was a small one. The first protein to be synthesized in the laboratory was oxytocin, a hormone with important functions in the body. Oxytocin is extremely small for a protein molecule: it consists of only eight amino acids. In 1953, the American biochemist Vincent du Vigneaud succeeded in synthesizing a peptide chain exactly like that thought to represent the oxytocin molecule. And, indeed, the synthetic peptide showed all the properties of the natural hormone. Du Vigneaud was awarded the Nobel Prize in chemistry in 1955.
More complicated protein-molecules were synthesized as the years passed; but in order to synthesize a specific molecule with particular amino acids arranged in a particular order, the string had to be threaded, so to speak, one at a time. That was as difficult in the 1950s as it had been a half-century earlier in Fischer’s time. Each time a particular amino acid was coupled to a chain, the new compound had to be separated from all the rest by tedious procedures, and then a new start had to be made to add one more particular amino acid. At each step, a good part of the material was lost in side reactions, and only small quantities of even simple chains were formed.
Beginning in 1959, however, a team under the leadership of the American biochemist Robert Bruce Merrifield, struck out in a new direction. An amino acid, the beginning of the desired chain, was bound to beads of polystyrene resin. These beads were insoluble in the liquid being used and could be separated from everything else by simple filtration. A new solution would be added containing the next amino acid, which would bind to the first. Again a filtration, then another. The steps between additions were so simple and quick that they could be automated with almost nothing lost. In 1965, the molecule of insulin was synthesized in this fashion; in 1969, it was the turn of the still longer chain of ribonuclease with all its 124 amino acids. Then, in 1970, the Chinese-American biochemist Cho Hao Li synthesized the 188-amino-acid chain of human-growth hormone. In principle, any protein can now be synthesized; it only requires that enough trouble be taken.
THE SHAPE OF THE PROTEIN MOLECULE
With the protein molecule understood, so to speak, as a string of amino acids, it became desirable to take a still more sophisticated view. What is the exact manner in which that amino acid chain bends and curves? What is the exact shape of the protein molecule?
Tackling this problem were the Austrian-English chemist Max Ferdinand Perutz and his English colleague John Cowdery Kendrew. Perutz took as his province hemoglobin, the oxygen-carrying protein of blood, containing something like 12,000 atoms. Kendrew took on myoglobin, a muscle protein similar in function to hemoglobin but only about a quarter the size. As their tool, they used X-ray diffraction studies.
Perutz used the device of combining thc protein molecules with a massive atom, such as that of gold or mercury, which was particularly efficient in diffracting X rays. Thus, he got clues that allowed him more accurately to deduce the structure of the molecule without the massive atom. By 1959, myoglobin, and then hemoglobin, the year after, fell into place. It became possible to prepare three-dimensional models in which every single atom could be located in what seemed very likely to be the correct place. In both cases, the protein structure was clearly based upon the helix. As a result, Perutz and Kendrew shared the Nobel Prize in chemistry in 1962.
There is reason to think that the three-dimensional structures worked out by the Perutz-Kendrew techniques are after all determined by the nature of the string of amino acids. The amino-acid string has, so to speak, natural crease points; and when they bend, certain interconnections inevitably take place and keep it properly folded. It is possible to determine what these folds and interconnections are by working out all the interatomic distances and the angles at which the connecting bonds are placed, but it is a tedious job indeed. Here, too, computers have been called in to help, and these have not only made the calculation but thrown the results on a screen.
What with one thing or another, the list of protein molecules whose shapes are known in three-dimensional detail is growing rapidly. Insulin, which started the new forays into molecular biology, had its three-dimensional shape worked out by the English biochemist Dorothy Crowfoot Hodgkin in 1969.
Enzymes
Useful consequences follow from the complexity and almost infinite variety of protein molecules. Proteins have a multitude of different functions to perform in living organisms.
One major function is to provide the structural framework of the body. Just as cellulose serves as the framework of plants, so fibrous proteins act in the same capacity for the complex animals. Spiders spin gossamer threads, and insect larvae spin cocoon threads of protein fibers. The scales of fish and reptiles are made up mainly of the protein keratin. Hair, feathers, horns, hoofs, claws, and fingernails—all merely modified scales—also contain keratin. Skin owes its strength and toughness to its high content of keratin. The internal supporting tissues—cartilage, ligaments, tendons, even the organic framework of bones—are made up largely of protein molecules, such as collagen and elastin. Muscle is made of a complex fibrous protein called actomyosin.
In all these cases, the protein fibers are more than a cellulose substitute. They are an improvement; they are stronger and more flexible. Cellulose will do to support a plant, which is not called on for any motion more complex than swaying with the wind. But protein fibers must be designed for the bending and flexing of the appendages of the body, for rapid motions and vibrations, and so on.
The fibers, however, are among the simplest of the proteins, in form as well as function. Most of the other proteins have more subtle and more complicated jobs to do.
To maintain life in all its aspects, numerous chemical reactions must proceed in the body. These must go on at high speed and in great variety, each reaction meshing with all the others, for it is not upon anyone reaction, but upon all together, that life’s smooth workings must depend. Moreover, all the reactions must proceed under the mildest of environments—without high temperatures, strong chemicals, or great pressures. The reactions must be under strict yet flexible control and must be constantly adjusted to the changing characteristics of the environment and the changing needs of the body. The undue slowing down, or speeding up, of even one reaction out of the many thousands would more or less seriously disorganize the body.
All this is made possible by protein molecules.
CATALYSIS
Toward the end of the eighteenth century, chemists, following the leadership of Lavoisier, began to study reactions in a quantitative way—in particular, to measure the rates at which chemical reactions proceed. They quickly noted that reaction rates can be changed drastically by comparatively minor changes in the environment. For instance, when Kirchhoff found that starch could be converted to sugar in the presence of acid, he noticed that while the acid greatly speeded up this reaction, it was not itself consumed in the process. Other such examples were soon discovered. The German chemist Johann Wolfgang Döbereiner found that finely divided platinum (called platinum black) encouraged the combination of hydrogen and oxygen to form water—a reaction that, without this help, could take place only at a high temperature. Döbereiner even designed a self-igniting lamp in which a jet of hydrogen, played upon a surface coated with platinum black, caught fire.
Because the “hastened reactions” were usually in the direction of breaking down a complex substance to a simpler one, Berzelius named the phenomenon catalysis (from Greek words essentially meaning “
break down”). Thus, platinum black came to be called a catalyst for the combination of hydrogen and oxygen, and acid a catalyst for the hydrolysis of starch to glucose.
Catalysis has proved of the greatest importance in industry. For instance, the best way of making sulfuric acid (the most important single inorganic chemical next to air, water, and, perhaps, salt) involves the burning of sulfur—first to sulfur dioxide (SO2), then to sulfur trioxide (SO3). The step from the dioxide to the trioxide would not proceed at more than a snail’s pace without the help of a catalyst such as platinum black. Finely divided nickel (which has replaced platinum black in most cases, because it is cheaper) and such compounds as copper chromite, vanadium pentoxide, ferric oxide, and manganese dioxide also are important catalysts. In fact, a great deal of the success of an industrial chemical process depends on finding just the right catalyst for the reaction involved. It was the discovery of a new type of catalyst by Ziegler that revolutionized the production of polymers.
How is it possible for a substance, sometimes present only in very small concentrations, to bring about large quantities of reaction without itself being changed?
Well, one kind of catalyst does in fact take part in the reaction, but in a cyclic fashion, so that it is continually restored to its original form. An example is vanadium pentoxide (V2O5), which can catalyze the change of sulfur dioxide to sulfur trioxide. Vanadium pentoxide passes on one of its oxygen atoms to SO2, forming SO3 and changing itself to vanadyl oxide (V2O4), But the vanadyl oxide rapidly reacts with oxygen in the air and is restored to V2O5. The vanadium pentoxide thus acts as a middleman, handing an oxygen atom to sulfur dioxide, taking another from the air, handing that to sulfur dioxide, and so on. The process is so rapid that a small quantity of vanadium pentoxide will suffice to bring about the conversion of large quantities of sulfur dioxide; and in the end, the vanadium pentoxide appears unchanged.
