EDMAN, PEHR VICTOR

(b. Stockholm, Sweden, 14 April 1916;

d. Munich, Germany, 19 March 1977), chemistry, protein structure, amino acid sequence.

In the history of protein chemistry Pehr Edman will be remembered for his work on amino acid sequences in proteins. He developed a technique by which amino acids in the protein molecule can be removed and identified one after the other in a stepwise fashion. The fast developments in molecular biology would not have been possible without this method, known as the Edman degradation, which is virtually the only tool for this purpose.

Edman set out already in the 1940s to solve the problem of rapid and accurate stepwise degradation of protein chains. Throughout his career he resisted all temptation to deviate from his set course until he considered the task completed and well done. Rewards for early application of his method to interesting biological and medical problems went to others. At the time of his death in 1977 he was still engaged in optimizing the method.

Origin and Early Education . Edman was born in Stockholm, Sweden, on 14 April 1916. His father, Victor Edman, a judge, was a serious man and a devoted Christian. His mother, Alba Edman, was joyous, lively, and neat. Edman attended public elementary school in Stockholm. He began high school with a focus on the humanities, but soon switched to a focus on mathematics and the natural sciences. In 1935 Edman passed his matriculation examination with excellent records.

Medical and Biochemical Studies . Edman began medical studies at Karolinska Institutet in Stockholm, Sweden. He received the bachelor of medicine degree in 1938 and graduated as a physician in 1946. Concurrently with his studies in medicine he trained in biochemistry under the guidance of Professor Erik Jorpes (famous for his studies on heparin; an inhibitor of blood coagulation). For a time he also studied in Professor Hugo Theorell’s department (Theorell received the Nobel Prize in 1955 for his studies on oxidative enzymes). He became interested in protein chemistry and started on a project of his own—isolation and characterization of angiotensin (also called hyper-tensin or angiotonin). One of his tools during the work was column chromatography, and, seeing the need for collection of fractions at short regular intervals, he invented the first automatic fraction collector. This work resulted in a thesis that was presented at the Karolinska Institutet in 1945, receiving the mark of summa cum laude. The thesis work took place during World War II, and for a time Edman was drafted to serve as a physician in the armed forces. After his dissertation, Edman was admitted as docent, or lecturer, at the Karolinska Institutet. In order to widen his experience and perspectives in protein chemistry he spent one year (1946–1947) at the Rockefeller Institute in Princeton, New Jersey (a division of the Rockefeller Institute in New York). The institute was, at the time, one of the most prestigious places for protein chemists. Two of its members, John Northrop and Wendell M. Stanley, received the Nobel Prize in 1946 for preparation of enzymes and virus proteins in pure form. Edman studied in the laboratory of Northrop and Moses Kunitz, the latter a master in preparing proteins in crystalline form.

Studies on Protein Structure . By 1950 the realization had dawned that individual proteins contained molecules that were identical replicas, had exact molecular masses and amino acid compositions, and had identical packing of the polypeptide chains. This view gained final acceptance when crystalline preparations of proteins became available for analysis of their x-ray diffraction pattern; it was found that crystals gave rise to diffraction patterns that could be interpreted on the same basis as the crystals of simple inorganic salt compounds, although this required tremendous computer capabilities.

Although isolation of proteins and the analysis of amino acid composition had essentially become routine, the sequence of the amino acids in the polypeptide chain was still unknown. The first successful attack on the problem of sequence determination in proteins was mounted by Frederick Sanger at the University of Cambridge, England, and published during the period 1945 to 1951. Sanger received the Nobel Prize in 1958 for his work on the structure of proteins, especially that of insulin, his model protein. Sanger’s idea for establishing the sequence was simple. First, all that is necessary for the determination of the sequence of a dipeptide is amino acid analysis and a method of determining which amino acid is N-terminal (that is, the amino acid having a free NH ₂ group) or which is C-terminal (having a free COOH group) (see Figure 1). This reasoning could be applied to

a protein with known N-terminal amino acid provided it could be broken down to smaller peptides by random partial acid hydrolysis or by digestion with proteolytic enzymes. After labeling the N-terminal amino acid of the isolated peptides with 1-fluoro-2,4-dinitrobenzene, the peptides were exposed to acid hydrolysis and the components of the hydrolysate identified by chromatography. The labeled N-terminal amino acid had a bright yellow color, facilitating its identification. By fitting overlapping short sequences extending from the N-terminal to the C-terminal end of the protein, the complete amino acid sequence could eventually be deduced.

There is now strong evidence that the biological activity of a protein is somehow linked to its amino acid sequence, that is, to its primary structure. Attachment of prosthetic groups (such as sulfur, phosphorus, lipids, and so forth) to the peptide chains as well as their folding into a tertiary structure are also important for biological activity, but both attachment and folding are dependent on the primary structure. This view reflects Edman’s thinking in the early 1960s. At that time it received experimental support from Chinese scientists, at the Academia Sinica in Shanghai, who synthesized insulin chains de novo from synthesized amino acids and showed that the compound formed had almost full biological insulin activity. During his earlier work on angiotensin it had become clear to Edman that molecular mass and amino acid compositions were insufficient to explain the biological activity of a protein. Most likely this must somehow reside in the amino acid sequence of the protein.

A New Idea Is Born . During his time at the Rockefeller Institute in 1946 and 1947, Edman made the first attempts toward solving the sequence problem using stepwise degradation of proteins by removing amino acids one after the other. The problem was how to break a terminal peptide bond while leaving the rest unaffected. He was convinced that the method used by Sanger would have severe limitations when it came to sequencing large protein molecules.

Whereas Sanger wanted a stable label on the N-terminal that could survive acid hydrolysis of all peptide bonds, Edman turned his attention to chemical reagents that would allow both labeling of the N-terminal amino acid residue in proteins and subsequent rearrangement and release of the N-terminal amino acid under conditions that would not lead to breakage of other peptide bonds. This approach was initially influenced by the work of previous investigators. Max Bergman and coworkers had in 1927 used phenylisocyanate to label the N-terminal amino acid in dipeptides and had shown that on subsequent acid hydrolysis the phenylhydantoin of the N-terminal was formed. In 1930, Emil Abderhalden and Hans Brockmann had been the first to use phenylisocyanate for stepwise degradation of polypep-tides. However, it was inconvenient for sequence purpose because the labeled N-terminal amino acid could not be released without breaking other peptide bonds. Edman wanted to find another, more efficient, nucleophile, that is, a label on the N-terminal that would present the CO-group of the linkage to the next amino acid with an energetically more favorable reaction partner (than the NH-group) and thus cause fission of the bond without breaking other peptide bonds. Edman considered phenylisothiocyanate to be such a nucleophile, because the ketonic sulphur in this compound would be expected to be a good supplier of electrons (see Figure 2-1).

On his return to Sweden in 1947, Edman was appointed associate professor at the University of Lund and there he continued to work on the use of phenylisothiocyanate for sequence analysis. He continued work on other problems in protein chemistry, but sequence analysis was his major scientific interest. In 1949 the first version of Edman’s method for determination of the amino acid sequence of peptides was ready for publication. It featured a concept of sequencing that was novel and which ultimately had far greater potential than Sanger’s method, both for development and also for automation. In this first method Edman showed that coupling of phenylisothiocyanate to amino groups of peptides and proteins occurred easily in slightly alkaline buffers (see Figure 2-2). Furthermore, the labeled N-terminal amino acid was, in anhydrous acid media, swiftly rearranged and released as, he believed, a “phenylthiohydantoin” derivative of the N-terminal amino acid (see Figure 2-5). Following investigations on the reaction mechanisms and procedures for characterization of the “thiohydantoins” of different amino acids, a second version of Edman’s method was published in 1950.

At the time of the above studies, Edman was unaware of an intermediate released from the labeled peptide in anhydrous acid media. He discovered this intermediate during a thorough study on the reaction mechanism of the thiohydantoin formation. The resulting publication in

1956 was of utmost importance for the practical execution of stepwise amino acid sequencing, as it now became evident that the product formed in anhydrous acid media was not a phenylthiohydantoin (PTH) derivative, as Edman had assumed before, but rather a novel compound, the isomeric anilinothiazolinone (2-anilino-5-thiazolinone) (see Figure 2-3). This derivative has different absorption spectrum in ultraviolet light and different migration during paper chromatography, which, presumably, is what initially led Edman to make a thorough chemical characterization of the compound. Its formation following the nucleophilic attack on the peptide bond by the thioketonic sulfur in phenylisothiocyanate was extremely fast. Its importance for the success of stepwise degradation lies in its speed of formation and in the fact that its formation is not due to a hydrolytic process but occurs under water-free acid conditions. The risk for cleavage of other peptide bonds than that involving the N-terminal amino acid is thus negligible. After its release it can be removed from the residual peptide by extraction and subsequently under hydrolytic acid conditions transformed to the thiohydantoin. The latter occurs in two steps. First the thiazolinone undergoes rearrangement in water to the corresponding thiocarbamyl derivative of the amino acid (see Figure 2-4) and then under hydrous acid conditions cyclization to the thiohydantoin takes place (see Figure 2-5). The latter reaction is slower than the thiazolinone formation. All PTH derivatives of amino acids show strong absorption in the ultraviolet (with a maximum around 268 nm), which is useful for their quantification.

It was now clear that three discrete reaction steps were involved in stepwise degradation as shown in Figure 2: first, labeling of N-terminal amino acid of peptide or protein with phenylisothiocyanate (1-2); second, cleavage to form anilinothiazolinone derivative under anhydrous acid conditions (3); followed by rearrangement to thiocarbamyl derivative of N-terminal amino acid (4); and, third, conversion to phenylthiohydantoin derivative under hydrous acid conditions (5).

The generality of thiazolinone formation in stepwise amino acid sequencing has been demonstrated in many other stepwise reactions using other reagents but where the key reaction is always the formation of a thiazolinone. Edman’s mother reflected on the importance to Edman of the discovery of the three-stage reaction: “One day Pehr came home to me and asked me to sit down with him because he had something interesting to tell me. He then told me that he had discovered a way to analyze proteins which had not been possible before and that this discovery would certainly be of great importance for biochemistry in the future” (personal recollection of the author).

Edman characterized PTH derivatives of most of the naturally occurring amino acids by melting point and elementary analysis. With a few exceptions the chemical stability of PTH amino acids was excellent. Edman’s coworker John Sjöquist developed paper chromatography systems for resolution of PTH amino acids. These systems permitted a direct identification and quantification of all amino acids as PTH derivatives instead of, as previously, indirectly by amino acid analysis after alkaline hydrolysis of the PTH derivatives. Later, a whole arsenal of procedures for rapid identification of the PTH amino acids was developed, for example, thin-layer and gas chromatography, mass spectroscopy, and high performance liquid chromatography.

Edman’s approach of successively removing and identifying the N-terminal amino acid from a long polypep-tide demanded that each step should be as precise and as free from side reactions as possible. This stepwise degradation must end as soon as the product of the side reactions reached concentrations comparable with those of the linear degradation. By eliminating side reactions and modifying the reaction mechanisms, Edman strived, up to the end of his life, to increase the repetitive yield of the N-terminal amino acid from one degradation cycle to the next. Edman illustrated the importance of high repetitive yields with a simple calculation: repetitive yields of 97, 98, and 99 percent make possible 30, 60, and 120 degradation cycles, respectively.

Some researchers modified Edman’s method so that identification of the N-terminal amino acid was not performed directly, but instead by comparison of the amino acid composition of acid hydrolysates of the peptide before and after release of the N-terminal amino acid. Edman repudiated this approach on the grounds that it relates only to the hydrolysate of the peptide and, furthermore, that certain amino acids such as asparagines, gluta-mine, and tryptophan could not be indentified.

In 1957 Edman accepted a position as director of research at the newly established St. Vincent’s School of Medical Research in Melbourne, Australia, where he remained for fifteen years and also became an Australian citizen. There, Edman perfected the manual three-stage degradation technique. Briefly, the manual three-stage method was performed as follows: protein or peptide is coupled with phenylisothiocyanate. After coupling, excess phenylisothiocyanate and by-products are removed by extraction with benzene and the water phase freeze-dried. The residue is extracted with ethyl acetate and dissolved in water-free trifluoroacetic acid, which leads to the release of the amino terminal amino acid as an anilinothiazolinone derivative. Trifluoroacetic acid has the advantage of being both a good catalyst for cyclization and a good solvent for proteins. From this solution the residual protein or peptide is precipitated with ethylene chloride and is now ready for the next degradation cycle. The ethylene chlo-ride phase containing the thiazolinone is evaporated and the residue taken up in diluted hydrochloric acid for conversion of the thiazolinone to thiohydantoin at elevated temperature.

Except for his nearest coworkers, very few in the scientific community knew about the three-stage method. Edman refused to publish it as it stood at that time. He thought it could still be improved. Margareta Blombäck and this author worked in Edman’s laboratory as visiting scientists in 1961 and tried the three-stage technique on a 16–amino acid residue peptide from fibrinogen, the clotting protein in blood. (Release of the peptide activates fibrinogen.) We were able to make a complete stepwise degradation with good yields up to the very last residues. These results impressed Edman and strengthened his conviction that longer peptides could be degraded in even better yield.

Automated Procedure . The need for automation was evident by 1962. Edman and his laboratory assistant Geoffrey Begg explored different possibilities to solve the automation problem. The problem was first to find a single physical process that could accommodate the various operations in the manual procedure. At the very start of this endeavor they conceived the idea of the spinning cylindrical cup, in which all reaction media were spread out as thin films on the vessel wall (see Figure 3). This established a large surface and accomplished the equivalent of rapid stirring. The rotating film containing the protein would be well suited for extraction with a solvent, because the extraction fluid is continuously fed at the bottom of the cup, glides over the surface of the protein film, and is subsequently removed in the upper part of the cup. The surface film is also well suited for carrying out drying and other procedures, and, furthermore, the whole degradation cycle could be programmed.

This instrument was designed to contain reservoirs to hold all the reagents required for the reaction cycles together with receivers for effluents and means for controlling reaction temperature. A system of feed tubes and automated valves was provided and programmed to supply the reagents and extraction solvents to the spinning cup in the correct order at preset time intervals. The process embraced the coupling step with formation of the phenylthiocarbamyl derivation of the protein and release of the N-terminal amino acid as anilinothiazolinone. The thiazolinones of each N-terminal amino acid were automatically transferred to tubes in a fraction collector and were then, in a separate operation, converted to the corresponding PTHs for identification by thin-layer chromatography or other suitable procedures. The steps in the sequenator were largely copies of the manual three-stage method, but for logistic or other reasons certain changes had to be made. For example, trifluoroacetic acid in the cleavage step was changed to heptafluorobutyric acid to avoid excessive evaporation.

The instrument enabled the degradation of even large polypeptides to an extent that had never before been possible. The degradation cycle proceeded at a rate of about fifteen cycles in twenty-four hours, as compared to one or two cycles per day with the manual technique. The three-stage sequence method had changed the strategy of sequence determination, and automation made it widely available. It was no longer necessary to begin by cleaving the protein backbone into many small peptides, since long direct sequences were possible. Later on Edman applied the process to apomyoglobin from the humpback whale and showed that it was possible to establish the sequence

of the first sixty amino acids from the N-terminal end. The value of a sequencing technique depends to a large extent on the length of the amino acid sequence that may be determined. Therefore an important factor is high repetitive yield, that is, the yield of amino acid calculated from one degradation cycle to the next. With apomyoglobin this yield was 98 percent. In 1967 the work on the automated sequence analysis was finished and published. Edman’s work became well known by the international science community at large and clarification of protein structures worldwide sped up tremendously. Up to that time most sequencing had been made using the earlier versions of Edman’s method or with the Sanger technique.

In 1972 Edman accepted a position as director of the Department of Protein Chemistry I of the Max-Planck-Institut für Biochemie in Martinsried, near Munich, West Germany. His laboratory was endowed with a wealth of hard-won experience in sequencing and was engaged in a number of sequence projects. Nevertheless, during the last years of his life Edman continued to work on improving repetitive yield in the sequence procedure, which he thought was crucial for fast elucidation of protein sequences. With the advance of molecular genetics, high repetitive yields may have lost some of its former imperative. Relatively short amino acid sequences of a protein are required for identifying the cDNA sequence corresponding to the protein, and that sequence is subsequently translated into a (virtual) amino acid sequence.

Another concern that Edman had at this time was the necessity for the establishment of adequate arrangements for data storage, data retrieval, and data processing when a large number of protein sequences became clarified. For Edman, one of the most potentially powerful applications of sequence studies lay in the search for evolutionary relationships between different proteins, but this becomes valuable only when a large number of sequences are available for comparison. In his own words, “we may in time expect the unraveling of a new systema naturalis among biomolecules” (personal recollection by the author). He also foresaw early that mutational events in primary structure could change the function of a protein. Vernon Ingram had already in 1958 shown that a single point mutation in hemoglobin gave rise to sickle-cell anemia. Likewise in 1968 Birger Blombäck and others demonstrated that a point mutation in fibrinogen led to a severe bleeding diathesis.

Though Edman’s most important work was concerned with stepwise degradation of proteins, he also worked in other areas of biochemistry. His work on hyper-tensin has already been mentioned. Early on he worked on constituents in nerve tissues. He later developed a technique for partition chromatography on starch and used it for separation of nucleic acid components. He experimented with the coupling of proteins to insoluble matrices. He studied the mechanism of cleavage of proteins by cyanogen bromide in order to obtain convenient fragments for stepwise degradation. He observed as early as the 1930s that some proteins such as fibrinogen had a lower solubility in the cold and consequently could be frozen out of solution. Much later, other workers were using this physical property for purification of fibrinogen, antihemophilic factor, and von Willebrand factor.

Edman preferred to work in a modest setting with only a few people around him. In his group in Lund, Sjöquist worked on a method for amino acid analysis using their PTH derivatives and Lars Josephsson studied reversible breakage of peptide bonds in anhydrous acid solution, the so-called N-O acyl shift. His group in Melbourne also consisted of a small number of people: Frank Morgan and Hugh Niall worked mainly on applications of the phenylisothiocyanate degradation technique. Derek Ilse studied the mechanism of the reaction. In Melbourne and during the time in Martinsried, Agnes Henschen and her group studied the structure of fibrinogen and were able to determine the larger part of the amino acid sequence of this protein with more than a thousand amino acid residues.

Edman received several honors for his achievements in science: the Britannica Australia Award, the Berzelius Gold Medal, the Gold Medal of the Swedish Academy of Engineering, and the Linderström-Lang Medal. He was a Fellow of the Australian Academy of Science, Fellow of the Royal Society of London, and a scientific member of the Max Planck Society.

Edman, the Person . People who met Edman for the first time may have gotten the impression of a courteous, kind, but reclusive man with a hint of shyness. People who came closer to him could appreciate other qualities: generosity, warmth, humor, sympathy. Edman had a vast knowledge in many areas. His mind was logical and he was stringent in expression. The integrity on which his opinions were based was admirable and respected. At the core of his personality was a sincere humanism.

Edman had an urge for purity and perfection in life and work. This quality was very likely in play when he joined the socialist group Clarté as a young man in the 1930s and when he chose self-imposed expatriation in the 1950s, and it was probably a strong driving force in his scientific accomplishments. This urge was most likely a prerequisite for his motivation to spend so much time and effort on perfection of the phenylisothiocyanate method. Edman was suspicious of fortuitous experiments for the simple reason that if successful they would be difficult to reproduce.

Pehr Edman had broad interests outside science. His love for nature derived from his childhood. He was especially interested in birds, and his knowledge of ornithology was impressive. Music, classical or contemporary, and literature were other favorite preoccupations.

Edman had many friends, many of them from circles outside the scientific field. Among them his shyness seemed to disappear and his humor blossomed—sometimes drastic but to the point. Edman’s first marriage with Barbro Bergström produced two children, Martin and Gudrun. Edman met Agnes Henschen in Melbourne in 1966, and they were married in 1968. They had two children, Carl and Helena. Henschen later moved to the United States and worked in the Department of Molecular Biology and Biochemistry at the University of California, Irvine.

In February 1977, when leaving a scientific lecture in Martinsried, Edman suddenly fell down, unconscious. After a few weeks of illness he died on 19 March. A tumor of the brain was diagnosed; there had been no symptoms before he was struck unconscious.

BIBLIOGRAPHY

WORKS BY EDMAN

“A Method for the Determination of the Amino Acid Sequences in Peptides.” Archives of Biochemistry and Biophysics 22 (1949): 475–490.

“Preparation of Phenylthiohydantoin from Some Natural Amino Acids.” Acta Chemica Scandinavica 4 (1950): 277–282.

“Method for Determination of the Amino Acid Sequence in Peptides.” Acta Chemica Scandinavica 4 (1950): 283–293.

“Selective Cleavage of Peptides.” In The Chemical Structure of Proteins, edited by G. E. W. Wolstenholme and Margaret P. Cameron. London: Churchill, 1953.

“Mechanism of the Phenyl Isothiocyanate Degradation of Peptides.” Nature 177 (1956): 667–668.

“On the Mechanism of the Phenyl Isothiocyanate Degradation of Peptides.” Acta Chemica Scandinavica 10 (1956): 761–768.

With John Sjöquist. “Identification and Semiquantitative Determination of 3-Phenyl 2-Thiohydantoins.” Acta Chemica Scandinavica 10 (1956): 1507–1509.

With K. Lauber. “Preparation of Phenylthiohydantoins from Glutamine, S-carboxymethylcysteine and Cysteic Acid.” Acta Chemica Scandinavica 10 (1956): 466–467.

With K. Heirwegh. “Purification and N-terminal Determination of Crystalline Pepsin.” Biochimica et Biophysica Acta 24 (1957): 219–220.

“Phenylthiohydantoins in Protein Analysis.” Annals of the New York Academy of Sciences88 (1960): 602–610.

“Determination of Amino Acid Sequences in Proteins.”Thrombosis et Diathesis Haemorrhagica, Supplementum13 (1963): 17–20.

With Geoffrey Begg. “A Protein Sequenator.” European Journal of Biochemistry 1 (1967): 80–91.

With Agnes Henschen. “Sequence Determination. In Protein Sequence Determination: A Sourcebook of Methods and Techniques, edited by Saul B. Needleman. 2nd rev. and enl. ed., 232–279. Berlin: Springer-Verlag, 1975.

“Unwinding the Protein.” Carlsberg Research Communications 42 (1977): 1–9.

OTHER SOURCES

Blombäck, Birger. “Pehr Victor Edman: The Solitary Genius.” In Comprehensive Biochemistry 42, edited by Giorgio Semenza and Anthony John Turner, chapter 3, 103–135. Amsterdam: Elsevier Science BV, 2003.

Partridge, S. Miles, and Birger Blombäck. “Pehr Victor Edman.” In Biographical Memoirs of Fellows of the Royal Society 25 (1979): 241–265.

Birger Blombäck