From Gene to Protein: Translation into Biotechnology

conference.

PREFACE

The Miami Winter Symposia are now well established on the national and international scene as one of the major annual expositions of the new biology of the 1970s and 1980s. This nineteenth volume is the record of the proceedings of the fourteenth symposium held in Miami Beach in January 1982. The theme was the translation of the new basic research findings into the practical application of biotechnology, with reviews of methodology and the applications of such methodology that lie behind the practical innovations.

The theme of the symposium was set by the Feodor Lynen Lecturer, Cèsar Milstein, whose development, with George Köhler, of monoclonal antibodies promises to be the dominating tool of medical technology in the next decade, both for diagnosis and therapy. The symposium began with reviews of techniques of eukaryotic cell culture, hybridoma technology and uses, and the in vitro synthesis of DNA and its use in the generation of protein analogs. Cloning into eukaryotic cells and methods of increasing the levels of gene expression were sessions that clearly reflected current areas of intensive research that have important commercial and clinical value. The formal presentations concluded with descriptions of the biological activities of cloned gene products, including reports on trials with human subjects of interferon, human insulin, and growth hormone, reports that indicate how far and how fast this field has moved in the space of less than a decade since the technologies were developed. A panel session on horizons in biotechnology concluded the meeting, with the speakers looking forward to the directions of future research and its applications.

The symposium drew a capacity audience of almost 800. We are gratified to have been so successful in our efforts to stage a timely and topical meeting, and we believe that this volume, which also includes the discussions of each presentation, will be a most useful source of reference to basic scientists and biotechnologists alike.

Our ability to organize these meetings and subsequently publish the proceedings depends heavily on the help we receive from our faculty colleagues. Special thanks are due to the committee chaired by Thomas R. Russell and the secretarial staff, notably Sandra Black, Olga Sanchez, and Pat Buchanan. We are also most indebted to Dr. Ralph W. F. Hardy for his convening of the panel on Horizons in Biotechnology.

The symposium was made possible in part by the financial assistance of Abbott Laboratories; Beckman Instruments, Inc.; Eli Lilly and Company; Hoffmann-La Roche, Inc.; ICN Pharmaceuticals, Inc.; Merck Sharp & Dohme Research Laboratories; New England Nuclear; Smith Kline & French Laboratories; and the University of Miami School of Medicine, Departments of Dermatology and Pathology, and the Office of the Dean. Our special thanks are due to Bethesda Research Laboratories, Inc., who have become the sponsors of the Feodor Lynen Lecture.

Fazal Ahmad, Julius Schultz, Eric E. Smith and William J. Whelan

The Thirteenth Lynen Lecture

MESSING ABOUT WITH ISOTOPES AND ENZYMES AND ANTIBODIES

César Milstein,     Medical Research Council, Laboratory of Molecular Biology, Cambridge, England

Publisher Summary

Monoclonal antibodies were an unexpected by-product of pure, basic, research. Sometimes, the contribution of the fundamental research is not so obvious and a good example is the automation of biochemical analysis in hospitals. With the emergence of multinational companies and huge industrial complexes, the research and development component in industry has spread into general aspects of science. The emergence of high technology areas is eroding the boundaries between applied and basic research. Industry interest in science remains narrow-minded as industry is motivated by profit rather than by social benefit so it will remain narrow-minded. Granting bodies themselves, as well as laboratories and scientific institutions, under the pressure of the economic recession, are running the risk of undermining the basis of their existence by trying to compete with industrial enterprise, public or private. It is true that administrators in charge of negotiating with the government for funds may find commercial success a way out or at least a persuasive argument. However, what is necessary is a firmer basis for the support of basic science, especially because the past has shown that this is money well invested.

Looking back over my life, everything seems to have happened by accident

A.J.P. Taylor

La oportunidad la pintan calva

Old Spanish Proverb

I never imagined one could do amino acid seqences by the changes in electrophoretic and chromatographic mobility of radioactive peptides until I was actually doing it myself. This was in 1960, I was following Fred Sanger’s instructions, and when we established the sequence of the pentapeptide involved in the phosphorylation site of phosphoglucomutase I began to really appreciate what it meant to have fun in research. And since the Lynen Lecture traditionally includes personal notes, I might as well start by telling you how it happened. This may seem pretty irrelevant to the subject of the symposium, and let me put you at ease, because it is totally irrelevant! Trouble is, in addition to being irrelevant, it is likely to be boring! The blame goes squarely on the organisers first for having decided on a lecture instead of something more entertaining, and second for choosing me as the lecturer. They should know better.

I took my degree in Argentina, at the University of Buenos Aires. I wasn’t a particularly good university student. My major preoccupation was not academic matters, but the Students’ Union, and the students’ involvement in political and social matters. But somehow I stumbled through to the end of my exams. By then, I was working part-time in a laboratory of clinical biochemistry, where I learned what it meant to work efficiently; organise your time, choose your methods carefully, pre-define the level of accuracy, and work as fast as is compatible with that level of accuracy.

MY ENZYME PERIOD

So I had to decide what I was going to do with my life and with my degree. I reasoned that the way to make money was to continue with clinical biochemistry. The alternative was to try and do science, but this was a truly romantic hope. My university training was poor, and strongly biased towards a rather snobbish and remote view of what science was. It was clear that if I wanted to have a go at science, I should not start unless I could find someone I could trust to be a scientist, and not a fake. More or less by accident, I discovered that there was a biochemist called Leloir. Leloir gave one of these Winter Symposia lectures, which was entitled I hate boring people with my recollections. I am sure we all hate people being bored by our recollections. I will go even further; I will say that I get furious when people get bored with my recollections.

But, going back to my story, I went to see Leloir, who could not take me, and suggested that I should go and see Stoppani. Although I had never heard of him, he was the new Professor of Biochemistry at the Medical School, and somehow I felt that I was along the right tracks. Stoppani said that he would take me as a research student, but that he could offer no economic prospects at all, either in the form of a fellowship, or of a future promise of any description. The mention of those points was more than I actually expected and the subject suggested, the role of -SH groups in aldehyde dehydrogenase, met with my complete approval. I had no idea of what it was all about, but it seemed to have the correct blend of chemistry and biology. And so that’s the way I started messing about with enzymes.

When I look back to that early period, when Stoppani was one of the few, and perhaps the only, full-time Professor of the Faculty of Medicine in the University of Buenos Aires, a full-time professor who probably had a salary of about the same order of magnitude as a janitor, trying to do serious and honest research in a laboratory with no funds at all, I must confess that the idea that we were messing about with enzymes seems today almost too pretentious.

My first failure, which almost cost me my position in the lab, was to break, successively, three five-litre round bottom flasks, out of a total of five. A fortune, one of the most precious pieces of equipment in the laboratory. The most precious piece of equipment was a Warburg apparatus, which Stoppani didn’t allow anybody to use but himself. My first success was the development of a workable and reproducible protocol for the preparation of yeast acetone powders from which the enzyme aldehyde dehydrogenase could be extracted in good yield. It was the Argentinian answer to the insurmountable problem of the extraction methods of the original paper which called for liquid nitrogen. We could afford neither the liquid nitrogen, nor indeed the appropriate thermos flask. Acetone and ether were all right, because they could be recovered by distillation. As you can see, I learned biochemistry the hard way.

The situation in Argentina changed quite considerably after 1955, and in a short time we were preparing our enzyme using a refrigerated centrifuge, and we could even assay activity, not by the old Warburg method but with a spectrophotometer, which we could borrow from a richer neighbouring laboratory. And then I really started getting involved in the kinetic and catalytic properties of active sites (1,2). By the time I wrote my thesis, I had a solid background of enzymology. I even had a proper job to come back to when I was awarded a fellowship from the British Council to go to work in Cambridge.

CAMBRIDGE - PART 1

In 1958, the Department of Biochemistry at Cambridge was a mixture of the old and the new. My research supervisor was Malcom Dixon, and only a few doors away from the enzyme unit, Dorothy Needham and Robin Hill were still active in the laboratory. But in his unassuming way, Sanger, who that year, only two weeks after my arrival at the laboratory, was awarded his first Nobel Prize, was a dominant influence. A high-voltage electrophoresis room had a warning at the entrance, where someone had altered the original D in the word Danger to read Sanger - High Power!

Dixon suggested that I work on the enzyme phosphoglucomutase, to clarify some odd observations made a number of years earlier at the department, that phosphoglucomutase required two metals for full activity, magnesium and a trivalent metal like chromium. Phosphoglucomutase had been the subject of two controversies at that time. The first had to do with the requirement of a co-enzyme for activity. The second involved activation by chromium. The first controversy was resolved in Argentina by Leloir and co-workers, who demonstrated that the true co-factor was glucose 1,6-diphosphate, which had been a previously unrecognised impurity. It was with some pleasure that I could sort out the second and much less important controversy to show that the activation by chromium and by metal chelators was of the same nature, in both cases involving the removal of heavy metals which were highly toxic to the enzyme (3).

I suspect that the choice of phosphoglucomutase on the part of Malcom Dixon had another element, which he didn’t mention much. Considerable interest in this enzyme had been generated by a report by Koshland and Erwin, that the enzyme contained an active serine in an amino acid sequence that was similar to that of the proteolytic enzymes (4). This serine was phosphorylated by glucose 1,6-diphosphate, and since the serine phosphate derivative was stable to acid hydrolysis, radioactivity could be easily used to follow the derived peptides.

Fred Sanger at that time was actually looking for the development of methods which would allow amino acid seqences to be determined, using radioactive techniques. One day, we happened to be having tea at the same table, and he asked me what I was doing. He vaguely knew of my existence, because on arrival at Cambridge, he had been extremely kind, and arranged for my wife Celia to work with Kenneth Bailey. When I said that I was working on phosphoglucomutase, he immediately asked me whether I was going to work on the active site. I was extremely reluctant to do so. The radioactive preparation I estimated would take me at least 3 weeks, and the prospect of just confirming the previous sequence wasn’t a very good incentive. But each time I went to Sanger’s laboratory to use the only pH meter which worked reasonably well in the whole department, Fred kept asking me when I was going to make a radioactive preparation. One day it dawned on me that perhaps I didn’t have to purify the [³²P]-labelled glucose 1,6-diphosphate. 1 simply had to make a crude preparation, which would only take a few hours, and then allow the purified enzyme inside a dialysis bag to equilibrate with the labelled substrate. Then by dialysing against saline, I would end up with a labelled enzyme preparation, without having to work at all. When I explained this project to Fred, I was rewarded with a grin of approval, and encouraging noises. It is possible that this idea was the one in my whole scientific career which had the biggest impact on my personal future.

A few days later, we were running an electrophoresis of a partial acid hydrolysate of the labelled enzyme. The pattern which emerged from the autoradiograph did not bear any resemblance to the pattern of the proteolytic enzymes. The amount of material we had was ample at the radioactive level, but utterly insufficient for purification and sequence analysis. And I started chasing radioactive spots to try to derive an amino acid sequence by new methods. Fred had some ideas on how to do this, and the active centre of phosphoglucomutase appeared to be the right challenge.

Although the work is largely forgotten, it is an example of how his mind was working around 1960, a couple of years before he started following radioactive spots derived from nucleic acids. The first step was to establish the correlation between the peptides (Fig. 1). So each band was eluted, and subjected again to partial hydrolysis. A band could give only phosphoserine, in which case it was a dipeptide, it could give a band identified as a dipeptide, plus phosphoserine and nothing else, in which case it would be a tripeptide, and so on. In this way all the radioactive spots present in the partial hydrolysate could be mapped as derived from a pentapeptide in which the phosphoserine was in the middle. Removal of the N-terminal residue by Edman degradation gave rise to a smaller peptide defining the polarity of the map.

FIGURE 1 The correlation between the (³²P)-peptides of a partial acid hydrolysate of (³²P)-phosphoglucomutase. Taken from (5).

The next step was to find out what the amino acids that made up the dipeptides and tripeptides were. Charged amino acids were easily spotted. Titration using mobility data showed that one of the amino acids next to the phosphoserine was histidine. We were excited about the presence of a histidine attached to a phosphoserine and we decided to test it by diagonal electrophoresis. Fred had been thinking about making use of diagonal electrophoresis to spot changes in mobility of peptides, while I was reading about the sensitivity of histidine to photo-oxidation in the presence of methylene blue. The two things came together on a Saturday morning when, if my memory serves me right, Fred and I were discussing with Richard Ambler this likely histidine. The idea was to spread the peptides on a first dimension, expose the paper to ultraviolet light in the presence of methylene blue, dry and then run again on a second dimension at the same pH to see whether the treatment had any effect on the mobility of peptides. The result was there; it was a beautiful diagonal with the peptides containing histidine moving away from the diagonal position (Fig. 2).

FIGURE 2 The identification of histidine next to (32)-SerP by diagonal electrophoresis. Taken from (5).

Using specific reactions of this type, and partition chromatography data, we came up with a pentapeptide sequence (5) which had nothing to do with the previously reported sequence, and which I had to defend in several seminars I gave in the United States on my way back to take up my position in Argentina as head of the newly-created Division de Biologia Molecular at the Instituto Nacional de Microbiologia.

THE ARGENTINIAN INTERLUDE

The Instituto Nacional de Microbiologia was a rather old institution created around the model of the Institut Pasteur. After going through a period of great neglect, it got a fresh lease of life when a new director, Ignacio Pirosky, was appointed. He obtained from the Government special concessions, which allowed him to appoint a large number of very young scientists to full-time key positions. This was a very bold and imaginative move, which immediately put him and the young people appointed in direct conflict with the sclerotic old guard. But Pirosky had strong backing from the Government, while it lasted, and in the meantime an atmosphere of great scientific excitement was developing among the young newly-appointed full-time scientists. The Division of Molecular Biology was essentially a place where fundamental research was to be done, and in the Division I could count on a group of extremely gifted people.

I started a programme on alkaline phosphatase and phosphoglyceromutase as possible candidates on which active sites around phosphoserine residues could be studied. But of course, it is a well-known fact that governments in Argentina don’t last very long, especially if they are civilian governments appointed by popular ballot. So, about a year after my arrival in Argentina, we had a military coup, and a new Minister of Public Health. Predictably, this new Minister dismissed Pirosky, blaming him for all sorts of things, and young Argentinians being what they are, we became very emotionally involved in the defence of the dismissed Director. I am still surprised (a) that the officers of the Association of Scientists lasted as long as they did (1 year), (b) that the Minister himself lasted as long as he did, since in between there were two or three more sub-military coups, and (c) that in all these dealings and political upheaval, we could do any science at all (6,7,8).

As it happened, the Chairman and the Secretary of the Association of Scientific Staff were members of my division. When they were dismissed for the most appallingly trivial and unimaginative reasons, and without consulting me, I myself resigned, and wrote a letter to Fred Sanger, asking for a job. To my delight, a reply came by return of post, and very shortly after, I was back in Cambridge.

A JUMP INTO ANTIBODIES FROM THE DISULPHIDE BRIDGES

I arrived back in Cambridge in the spring of 1963. The Laboratory of Molecular Biology had already been functioning for just over a year (Fig. 3). When Fred suggested that I might start doing some experiments with antibodies, I quickly became interested. He suggested labelling tyrosines in the active centre of antibodies with radioactive iodine to study their amino acid sequences, as we had been doing with enzymes. That suited me very well, not only because I liked the idea of iodination of tyrosines, but also becuse in Argentina, I had been toying with the idea of reducing disulphide bonds of antibodies, labelling them with radioactive iodoacetate, and comparing the sequences around the labelled cysteine residues. The idea was to find out whether two different antibodies differed in primary structure, and, if they did, what the difference was. The critical experiment, therefore, looked deceptively simple.

FIGURE 3 The MRC Laboratory of Molecular Biology as it was in 1963.

After dozens and dozens of autoradiographs of peptide maps of iodinated DNP antibodies and normal immunoglobulins, I became convinced that this was a blind alley. A year ago, I disposed of those fingerprints, in a nostalgic but ruthless clear-up when I moved from one lab to another.

Fortunately, at the same time as I was doing these iodination experiments, I was also labelling the cysteines with ¹⁴C-iodoacetate. In the laboratory just across the corridor, the diagonal electrophoresis approach was being applied by Brown and Hartley, to define the disulphide bonds in proteins. Too near not to be attracted by the idea of applying the approach to γ-globulin and antibodies.

My ignorance of immunology was absolute. I was totally unaware of the vast literature already around at the time concerning the relations between myeloma proteins and normal immunoglobulins. So, when I noticed a very beautiful difference in the disulphide bridge diagonals of the light chains of a myeloma and a macroglobulin, I jumped with excitement, and rushed to London to tell Sidney Cohen and Rodney Porter that I thought I had discovered the difference between macroglobulins and γ-globulins. Sidney dampened my enthusiasm by saying That macro I gave you is type II. Before you jump to conclusions, you should try another myeloma protein, type I.

Well, this was news to me. What were type I and type II? Most of you probably don’t know what type I and type II are either, because the nomenclature has been totally dropped, but you will be on familiar ground if I tell you that type I and type II are what we now know as proteins containing κ and λ light chains respectively. So, obligingly, he gave me a little bit of another myeloma protein, this time 7S IgG, which was type I, ie with κ chains as opposed to the λ chains of the macroglobulin. Of course, that was the difference between the two fingerprints.

By a comparison between normal light chains and those coming from myeloma patients, I was able to build up a picture of disulphide bridges of the light chain which, published in 1964 (9), represented the first batch of sequence data on Bence Jones proteins. It came out just in time for me to be rather belatedly invited to the historic antibody workshop held in Warner Springs in the early part of 1965, where Hilschmann described for the first time the block sequence difference between two Bence Jones proteins demonstrating a constant and a variable part (10). My description in the 1964 paper of three disulphide bonds, one at the C-terminus being interchain, one being common to all Kappa Bence Jones proteins and light chains, and the other variable, fitted beautifully with the concept of the invariant C-terminus with one interchain and one intrachain disulphide bond and a variable N-terminus with another intrachain bond. And so, almost without realising it, I became an active member of the small bunch of the newly emerging molecular immunologists trying to understand the molecular nature of antibody diversity.

My studies of Bence Jones proteins culminated with the recognition of three subgroups of human kappa chains (11,12). Each subgroup (Fig. 4) contained a large family of chains, where the individuality of each chain was defined by amino acid sequence differences scattered along the chain, but mainly concentrated in certain regions, which are now recognised as the hypervariable regions. This also implied multiple genes for the V region. The postulate of Dreyer and Bennet (13) that V and C genes were separately encoded now seemed inescapable.

FIGURE 4 The three basic sequences of human Kappa chains.

With Richard Pink, my first research student in Cambridge, and then with Frangione, we attacked the S-S bonds of heavy chains. We became fascinated by the fact that the four subclasses of human IgG appeared to have derived from a common recent ancestor. That meant, for instance, that the γ1 immunoglobulin heavy chain of the human was not the homologue of the γ1 of the mouse. In this way, we became aware of a fundamental property of evolution in multigenic families, namely that the individual components were not in constant expansion, but rather in a continuous dynamic expansion and contraction process. Thus we suggested that the divergence point for the heavy chain classes was old in evolution, but the subclass evolution, being a much more recent event, did not have a common departure point for all the species. It was immediately obvious that, if one applied the same principles to the V regions, many of the puzzling aspects we had observed could easily be explained (14). Let me quote our conclusions from our 1970 paper: We suggest that the section of the genome involved in the coding of immunoglobulin chains undergoes an expansion-contraction evolution: that the number of individual genes coding for basic sequences is not large, and that it varies in different species and even within species at different stages of its own history. The task of providing for the endless variety of individual chains is left to somatic processes.

By then, the Division of Protein Chemistry had become the Division of Protein and Nucleic Acid Chemistry, and not surprisingly, I started to think about light chain messenger RNA. My first attempts were made much too early, in collaboration with Peter Fellner, a research student of Fred Sanger’s. Although the results were a clear disaster, with a hint of hope, the interest they created in my mind was strongly reinforced when George Brownlee felt that his success with the 6S RNA, the longest RNA sequence at that time, qualified him for more ambitious things. We worked frantically on solid tumours for over a year, but the results this time were very disappointing. We decided that we ought to move away from solid tumours, and work with tissue culture cells. But moving into tissue culture was not an easy decision. It all seemed too much like witchcraft.

George Brownlee and I started to grow mouse myeloma cells in culture which we imported from the Salk Institute. We had two or three to choose from, and the choice was based on easy growth on the one hand, and high production of immunoglobulin on the other. To assess production, we added to the culture medium a labelled amino acid, and then took the supernatant and subjected it to cellulose acetate electrophoresis. It was reassuring to look at a radioactive band, and the intensity of the radioactive band gave us a very good idea of the synthetic capacity of the individual cells. I must now resist the temptation to expand on my collaboration with Brownlee and later with Harrison and others, which led to the discovery of the precursor of light chains and its significance in secretion (15,16), and the sequence analysis of light chain mRNA (17,18).

Somehow the culturing of myeloma cells, our continued interest in somatic mutation and the potential of radioactive methods clicked together when a new research student, David Secher, was struggling with the sequence of a human myeloma protein, and feeling a bit unhappy about it. On an exciting Saturday morning, which was prolonged after lunch, David and I decided to look for somatic mutants of cells in culture. Only later were we to discover that Scharff and his collaborators were already doing excellent work along those lines.

We would have to learn how to clone individual cells to be able to screen for the large number of clones which would be required for those experiments. The fact that we were total beginners with tissue culture methods did not deter us. By extraordinary luck, we received reinforcement in the form of an Australian postdoc, Dick Cotton, who fortunately had no experience in tissue culture either. But he was willing to learn the little that we knew, and to carry on from there. And we embarked on a search for mutants of cells in culture. We settled for a protocol which is described in Fig. 5. The experiments were indeed very successful, and we did manage to obtain mutants (19). We then proceeded to sequence them, and to understand the chemistry of the mutations (20).

FIGURE 5 The protocol used for the screening of 7,000 clones of P3 myeloma cells. A number of structural mutants were detected, and are described in Table 1. Taken from