Stein, William Howard
STEIN, WILLIAM HOWARD
Stein was the second of the three children of Fred M. and Beatrice Borg Stein. His father was a businessman who retired early to devote full time to the New York Tuberculosis and Health Association, Montefiore Hospital, and other organizations concerned with the health of the community. His mother was also involved in community activities and was a pioneer in creating programs of afterschool activities for children.
Stein’s elementary education was at the Lincoln School, run by Teachers College of Columbia University. At the age of sixteen he transferred to Phillips Exeter Academy, a distinguished school with a demanding college preparatory program. In 1929 he entered Harvard College, following his father and older brother. After graduating in 1933 with a major in chemistry, Stein spent an additional year at Harvard to further his training in organic chemistry. In 1934 he began graduate studies in the department of biological chemistry at the College of Physicians and Surgeons of Columbia University, at that time one of the leading departments in the field; it had an outstanding faculty and a gifted assembly of graduate students, many of whom later became leaders in biochemical research.
In 1937 Stein completed his Ph.D. dissertation, on the amino acid composition of elastin, under the supervision of Edgar G. Miller, Jr. It was this initial foray into purifying and determining the structure of elastin that formed his lifelong devotion to the study of protein chemistry. In his doctoral investigations Stein used two new precipitating agents for isolating amino acids—potassium trioxalatochromate for glycine and ammonium rhodanilate for proline—that had recently been introduced by Max Bergmann.
It was thus natural that Stein decided to pursue postdoctoral research with Bergmann, who in 1934 had come to the United States from Dresden as a refugee from Nazi Germany and was on the staff of the Rockefeller Institute for Medical Research (now Rockefeller University). Bergmann had worked with Emil Fischer, one of the great pioneers in protein chemistry, and had made many important contributions to protein research before coming to the United States. He was particularly known for his development with Leonidas Zervas of the carbobenzoxy method for peptide synthesis.
In 1937, when Stein came to the Rockefeller Institute, Bergmann’s laboratory was one of the few in the United States devoted entirely to investigations of the chemistry of proteins and peptides, and to the specificity of proteolytic enzymes in hydrolyzing synthetic peptides and derivatives. As a result, over the years it attracted an outstanding group of postdoctoral fellows. When Stein joined the laboratory, he decided to investigate the use of the new precipitating agents that were being developed by Bergmann to improve quantitative gravimetric analysis of amino acids in protein hydrolysates. Stein introduced the solubility product method as a way of attempting to obtain more precise results, inasmuch as most precipitating reagents left a considerable amount of the amino acid in the mother liquor. In 1939, when Stanford Moore joined the laboratory as a postdoctoral fellow, Bergmann suggested that Stein and Moore combine their efforts to try to develop practical analytical procedures. During the next few years they were successful in developing precise methods for the analysis of glycine and leucine, two amino acids for which there had been no satisfactory analytical methods earlier. Thus began a collaboration that, except for an interval during the war years, lasted more than forty years.
In 1942, with the United States at war, research on amino acid analysis was stopped. The efforts of Bergmann’s laboratory were devoted to the study of the physiological and chemical actions of mustard gas and the nitrogen mustard compounds, in the hope of finding therapeutic agents to counter the effects of these highly toxic substances. Fortunately, none of these were used during World War II. Stein was a co-author of a series of papers, published after the war, on the chemical reactions of these compounds with amino acids and peptides. These studies provided valuable experience in understanding the reactivity of amino acid side chains. At the same time Moore served with the Office of Scientific Research and Development in Washington and elsewhere for some three years.
Max Bergmann died in 1944. Stein continued in the laboratory, where Moore rejoined him in 1945. Herbert Gasser, then director of the Rockefeller Institute, offered them the opportunity to initiate a joint program in protein chemistry, with the first efforts to be devoted to development of methods for precise amino acid analysis, an essential prerequisite for any detailed study of the structure of proteins.
In the meantime there had been a renaissance in separation methods by chromatography, largely because of the work performed during the war by Archer J. P. Martin and Richard L. M. Synge in England, and the development of liquid-liquid countercurrent distribution by Lyman C. Craig at the Rockefeller Institute. These achievements brought attention to the high resolving power of multiplate separations. Stein and Moore then began to develop chromatographic methods for separation of all of the amino acids in a protein hydrolysate. Their first efforts were concerned with the use of potato starch as a supporting matrix for partition chromatography with n-butanol-water as the resolving system, a procedure already attempted by S. R. Elsden and Synge, but with limited success.
After several years of intensive effort, elution systems with starch columns were developed for the quantitative determination of all the common amino acids in protein hydrolysates. These procedures, involving the use of three separate columns, were successfully applied to the analysis of bovine β-lactoglobulin and bovine serum albumin in 1949. The analyses required less than five milligrams of protein with a standard error of less than 5 percent, a remarkable achievement for that time.
The success of these analytical procedures was due to several significant advances in methodology. First, the metal contaminants in the potato starch had to be removed by use of 8-hydroxyquinoline in order to give reproducible results for the elution of each amino acid. Second, a photoelectric drop-counting method was developed to collect equal volumes of eluate from the columns. This, together with the invention of an automatic fraction collector, facilitated the process of collection. The machine that Moore and Stein developed became the prototype of similar commercial instruments that soon were used for all types of chromatographic procedures. Third, in order to estimate quantitatively the amount of amino acid in each test tube, a reagent form of ninhydrin was devised that produced a stable color with each eluted amino acid. This novel reagent, stabilized with a reducing agent and in an organic solvent miscible with water, was used to monitor the procedure. It was important that the developed color was proportional to the amount of amino acid in each fraction.
Although the starch-column chromatography was useful and represented an important advance, the procedures were slow, requiring a few weeks for a single complete analysis. The method was soon abandoned in favor of ion-exchange chromatography. Sulfonated polystyrene resins of reproducible properties in a finely powdered form had become available at that time. As a result, one or two columns with buffers at different pH values could be used in a procedure that shortened the elution time appreciably; these methods were published in 1951.
The methods were later made completely automatic by continuous elution at high pressure. The automatic machines, described in 1958, developed in collaboration with Darrel H. Spackman, gave a continuous flow of eluate from the column with the ninhydrin color produced in the stream of liquid. The optical density was recorded potentiometrically on continuous charts. A variety of commercial machines soon became available that had further practical advances in speed, sensitivity, and automation. Thus, within a relatively short time the ability to analyze the amino acids in protein hydrolysates, in body fluids, and in other biological materials had become routine and quantitative, readily available to every investigator.
Concurrently, beginning in 1949, Stein and Moore began to apply their chromatographic methods to the isolation and then to the study of the structure of bovine pancreatic ribonuclease. With the collaboration of a group of younger investigators (C. H. Werner Hirs, J. Leggett Bailey, Darrel H. Spackman, and Derek G. Smyth), the complete sequence was described between 1958 and 1960, the first enzyme for which the primary structure was established. The methods that were employed included cleavage of the oxidized peptide chain by various proteolytic enzymes or by cyanogen bromide, followed by separation of each of the resulting peptides by chromatography on ion-exchange columns with volatile buffers as the eluting agents, The sequence of each of the peptides was determined mainly by the repetitive application of the phenylthiohydantoin method developed by Pehr Edman for identification of the amino-terminal residue simultaneously with the removal of that residue.
With Arthur Crestfield and Moore, in 1963 Stein showed that iodoacetate, a known inhibitor, could react with either of two specific histidine residues, yielding a carboxymethyl group on the 1-nitrogen of histidine-119 or on the 3-nitrogen of histidine-12, but not with both on the same polypeptide chain. With Robert Heinrikson and Moore, Stein showed that carboxymethylation of lysine-41 inhibited the enzyme. From these data it was concluded that the reactive nitrogens of the two histidine residues were approximately five Ångstrom units apart at the active site of the enzyme and that the ε amino group of lysine-41 was seven to ten Ångstrom units from the nitrogen-3 of histidine-12. These predictions, based on purely chemical information, were later confirmed by X-ray crystallographic analysis of the three-dimensional structure elucidated independently by Frederic M. Richards and Harold Wyckoff at Yale and by G. Kartha, J. Bello, and David Harker in 1967.
Two further points on the studies of ribonuclease should be noted. First, Stein, Moore, and George Stark showed that carboxymethylation of the enzyme does not occur when it is unfolded in solutions of eight-molar urea. Thus, the specific reactivity of the two histidine nitrogen atoms depends on their proximity in the native enzyme, although one of the two histidine residues is near the amino-terminal end and the other is near the carboxyl-terminal end of the polypeptide chain of 124 amino acid residues. A by-product of this study was the realization that traces of cyanate in the urea solutions had carbamylated the ε amino groups of the enzyme, thus inactivating the ribonuclease. It is noteworthy that in 1828 Friedrich Wöhler had discovered the reverse reaction—that ammonium cyanate forms urea—a discovery that laid to rest the notion that organic compounds could be formed only in living organisms.
Ribonuclease forms active dimers when it is freeze-dried from a 50 percent solution of acetic acid. With Arthur Crestfield, Stein and Moore demonstrated that dimers also form when an equal mixture of the inactive derivatives 1-carboxymethyl-His-119 and 3-carboxymethyl-His-12 ribonucleases is freeze-dried in the same manner. About 25 percent of the mixture of dimers possesses half the activity of the native enzyme. They concluded that three kinds of dimers are formed; two of them have either alkylated His-119 or alkylated His-12 at the active sites, and are inactive. However, a third dimer is formed with one productive active site and one inactive site that contains both alkylated residues. Hence, such a dimer posseses only half the normal activity. This activity is lost when the dimers are dissociated to monomers. These experiments provided the first example of an active site of an enzyme that is formed by the noncovalent binding of two polypeptide chains, a situation later encountered by other investigators of several native enzymes.
Coincident with the work of Stein and Moore and their associates, Christian B. Anfinsen and his coworkers were investigating the properties of ribonuctease at the National Institutes of Health in Bethesda, Maryland. The contributions of both laboratories were recognized by the award of the Nobel Prize for chemistry in 1972 jointly to Stein and Moore and to Anfinsen. The Nobel Lecture by Moore and Stein was devoted mainly to the chemical studies on ribonuclease and deoxyribonuclease.
After the major studies on ribonuclease with 124 residues were essentially completed, Stein and Moore began work on the much larger pancreatic deoxyribonuclease, a glycoprotein with 257 residues in a single polypeptide chain. The sequence was established in 1973 with the collaboration of several students and postdoctoral fellows—Paul Price, Tehyung Liu, Brian Catley, Johann Salnikow, and Tahsiu Liao—and in later studies with Tony Hugli, Bryce Plapp, and Dalton Wang. In the purification of this enzyme, three chromatographically distinct active forms were isolated and identified. One of the forms, in minor amount, proved to contain a residue of proline for histidine at residue 118; since this form of the enzyme is fully active, His-118 is not essential. The carbohydrate side chain is attached via an aspartamidohexosamine linkage at residue 18; another active form of the enzyme possesses a shorter carbohydrate chain at the same residue. The enzyme is inactivated by nitration of tyrosine-62 and by carboxymethylation of histidine-131. In these studies, as in many others from this laboratory, the results were obtained by chromatographic isolation of highly purified fragments and careful quantitative analyses.
After the methods of ion-exchange chromatography for amino acid analysis and peptide separation were developed, they were applied to joint studies with other investigators at Rockefeller University on a variety of problems. In addition, a large number of other investigations were pursued, partly with the aim of training the many individuals who came to the laboratory to gain experience in the use of these methods. These studies included many on structure-function relationships in enzymes other than pancreatic ribonuclease and deoxyribonuclease. Among these investigations were those on bromelains (with Shoshi Ota), chymotrypsin (with Denis C. Shaw), pepsin (with T. G. Rajagopalan, T. A. A. Dopheide, and Roger Lundblad), streptococcal proteinase (with Teh-yung Liu, William Ferdinand, Brenda Gerwin, Norbert Neumann, Michael C. Lin, Michael Bustin, and Stuart D. Elliott), ribonuclease T1 (with Kenji Takahashi), 2′, 3′-cyclic nucleotide 3′ phosphohydrolase from the brain (with Arabinda Guha, David C. Sogin, and Robert J. Drummond), and carboxypeptidase Y (with Rikimaru Hayashi).
Concurrently with the investigations on protein structure, the methods of analysis by ion-exchange chromatography for amino acids and derivatives were modified for application to analyses of urine, blood, and extracts of animal tissues. From several studies of this sort (with Harris Tallan) came the identification of 3-methylhistidine and tyrosine-O-sulfate as normal constituents of human urine, the discovery that acetylaspartic acid and cystathionine are major constituents of human brain tissue, and the observation (with Alejandro C. Paladini and C. H. W. Hirs) that phenylacetylglutamine is a normal major metabolic product in human urine. In addition, quantitative estimates were made for many other previously known constituents of these fluids and of brain tissue. These results provided a framework for subsequent studies of this type in many laboratories devoted to investigations of human and animal metabolism, and of many clinically important metabolic disorders.
In all of the above studies, Stein and Moore continued their remarkable collaboration, one of the longest in any scientific field. It is impossible to judge which of the two was responsible for specific contributions to the work that emanated from their laboratory. Indeed, in Moore’s biographical memoir of Stein, he comments: “During the early years of our cooperation, Stein and I worked out a system of collaboration which lasted for a lifetime. Stein combined an inventive mind and a deep dedication to science with great generosity. Over a period of forty years, we approached problems with somewhat different perspectives and then focused our thoughts on the common aim. If I did not think of something, he was likely to, and vice versa, and this process of frequent interchange of ideas accelerated progress in research.”
Stein served on the editorial committee of the American Society of Biological Chemists from 1955 to 1961. He joined the editorial board of the Journal of Biological Chemistry in 1962, became an associate editor in 1964, and was named editor in 1968. He thus served during a period of great growth of the science and helped to further the progress of the Journal as one of the leading publications in the field.
In the summer of 1969, while attending an international symposium on proteolytic enzymes in Copenhagen. Stein came down with a high fever. A few days later, on the way home, he became paralyzed; the disease was later diagnosed as a severe case of Guillain-Barré syndrome. He survived the acute phase of the disease but, after a year of hospitalization, he remained a quadriplegic. Despite this handicap, for eleven years he maintained his interest in science and in the activities of the laboratory. He continued to consult with his collaborators in the laboratory and to hold informal seminars in his home on current research. In all of these efforts, he was firmly supported and encouraged by his wife, Phoebe Hockstader Stein, whom he had married in 1936, and by their three sons. Stein died suddenly of heart failure at his home in New York on 2 February 1980.
In addition to the Nobel Prize, Stein received many honors and awards. He was elected to both the National Academy of Sciences and the American Academy of Arts and Sciences in 1960. With Moore he shared the American Chemical Society Award in Chromatography and Electrophoresis in 1964, the Richards Medal of the American Chemical Society in 1972, and the Kaj Linderstrøm-Lang Award in 1972. He received honorary degrees from Columbia University and the Albert Einstein College of Medicine of Yeshiva University in 1973.
Stein spent almost his entire scientific career after his doctoral training at what is now Rockefeller University. He was away for brief periods as a visiting professor at the University of Chicago in 1961 and at Harvard University in 1964. His lectures to graduate students at Rockefeller, on visits to other institutions, and at scientific meetings were characterized by the same precision and rigor as his scientific research, and conveyed an enthusiasm for the developments in biochemistry that was undiminished to the end of his life.
It is useful to recall the large changes in the field of protein chemistry that developed as a result of the pioneering work of Stein and Moore. Studies by many investigators had firmly established the general biological importance and functional properties of proteins between the years 1925 and 1940. It was already known that aside from their structural roles, proteins could be enzymes, antibodies, hormones, oxygen carriers, and a host of other things, yet little was known of their structure. Not even the exact amino acid composition could be given for any protein. It was this gap in information to which Stein and Moore first devoted their investigations, in following up their fledgling efforts in Bergmann’s laboratory before the war. Their successful development of analytical procedures permitted a comprehensive attack on protein structure.
A great impetus for these efforts was provided by the work at Cambridge of Frederick Sanger, who with his co-workers was able to determine by 1952 that the hormone insulin, a small protein with two chains of thirty and twenty-one residues, has a unique amino acid sequence in a given species of animal. By 1955, the complete structure of insulin with its disulfide bridges had been ascertained. The methods used by Sanger were satisfactory for relatively small proteins but clearly would not be universally applicable for the bulk of proteins, most of them possessing polypeptide chains longer by at least an order of magnitude.
The work of Stein and Moore and their co-workers provided methods that, at least in theory, could be applied to proteins of almost any size and complexity. In itself, this provided a stimulus to the entire field. Part of the reason why their methods were adopted so rapidly and widely was the care they took to make the procedures easily reproducible even by novices. Further, they were careful to examine each minor artifact and difficulty that was encountered in order to resolve the problem. It was this meticulous attention to detail for each chemical and analytical procedure that rendered their methods so readily applicable to a host of problems in peptide and protein chemistry.
Stein and Moore’s methods have been complemented by the analytical procedures for determining the sequences of the deoxyribonucleic acids (DNA) of the individual genes that determine the sequence of the polypeptide chain of a protein. For very large proteins and for those that can be isolated only in minute amounts, the methods for determining the DNA sequence coding for the protein sequence are invaluable. Nevertheless, for study of the protein itself and its reactive residues, binding sites, and so on, the pioneering methods of Stein and Moore will continue to be essential. Indeed, it was their studies on the enzyme pancreatic ribonuclease that demonstrated that detailed understanding of the mechanism of action of such catalysts could be achieved purely by the use of chemical methods, once the amino acid sequence was determined. The meticulous investigations on the reactions of inhibitors with the enzyme produced a good picture of the nature and properties of the “active site.” Later, physical studies by various methods, including nuclear magnetic resonance and X-ray crystallography, served to confirm and expand knowledge of the properties of the amino acid side chains in the active site.
I. Original Works. A bibliography of Stein’s works is in Moore’s 1987 memoir (see below). Stein and Moore’s Nobel speech. “Chemical Structures of Pancreatic Ribonuclease and Deoxyribonuclease,” is in Les prix Nobel en 1972 (Stockholm, 1973), 120–143, and in Science, 180 (1973), 458–464.
II. Secondary Literature. On Stein, on Moore, and on their work, see Joseph S. Fruton, Molecules and Life: Historical Essays on the Interplay of Chemistry and Biology (New York, 1972), 87–179; Stanford Moore, “William H. Stein, in Journal of Biological Chemistry, 255 (1980), 9517–9518, “Dedication to William H. Stein,” in Chemical Synthesis and Sequencing of Peptides and Proteins: Proceedings of the International Conference Held 8–9 May 1980, at the National Institutes of Health, Bethesda, Maryland (New York, 1981), and “William H. Stein,” in Biographical Memoirs, National Academy of Sciences, 56 (1987), 414–440; Frederic M. Richards, “The 1972 Nobel Prize in Chemistry, in Science, 178 (1972), 492–493; and Emil L. Smith and C. H. W. Hirs, “Stanford Moore,” in Biographical Memoirs, National Academy of Sciences, 56 (1987), 354–385.
Emil L. Smith