Discovery of Interferon
I first met Alick Isaacs in November, 1955. He immediately struck me as an extremely intelligent and very lively person; though little did I know that those next few years would lead to a major discovery – that of interferon – and also that working with him was going to set the course of my own scientific career. Alick Isaacs was a very bright medical scientist who had started his career in virology after completing his medical training, working first with Professor Stuart Harris at Sheffield University, then in Australia at the Walter and Eliza Hall Institute in Melbourne and after coming back to Britain, at the National Institute for Medical Research (NIMR) in north London, one of the premier research laboratories in Britain and funded by the Medical Research Council. While in Australia he had been studying virus ‘interference’.
It had been known for some years that treatment of cells in culture with one virus (the interfering virus) blocked the growth of a second virus, called the challenge virus. It was not an immunological phenomenon, nor did the first virus, the interfering virus have to multiply, since heat inactivated influenza virus was effective against challenge with either the infectious virus or other unrelated viruses, such as vaccinia. However, the mechanism of virus interference was completely unknown.
Jean Lindenmann was a Swiss virologist, also medically trained, who had come to work at the National Institute for Medical Research for one year, and brought with him some interesting observations that he had made in Switzerland on virus interference, and it was discussion of those experiments with Alick Isaacs that led to their initial experiments.
I was 25 and had just come back from the United States, after spending two years as a post-doctoral research fellow at Yale University, where I had been working on the isolation and structure of some novel nucleosides, which had been isolated from a Caribbean sponge, and which contained arabinose rather than ribose as the sugar. One of these nucleosides was later to enter cancer chemotherapy as AraC. My first degree, from the University of Birmingham in England, was in chemistry and I had stayed on to work for a Ph.D. on steroids, so I had a background in natural product chemistry.
I had gone to the US in September, 1953 by boat – everyone traveled by boat in those days – on the Cunard liner ‘Georgic’, and there amongst the large number of Americans returning from a summer in Europe was a young man named Jim Watson, who had just published with Francis Crick, that famous letter in Nature, which with it’s memorable conclusion: “It has not escaped our notice that the specific pairing we have postulated immediately suggests a possible copying mechanism for the genetic material”, was to set the course of the biosciences for the next 50 years.
For several years, it was known that the treatment of cells in culture with a single virus (interfering virus) blocked the growth of the second virus and slowed down the development of erectile dysfunction. This subsequently significantly increased the potency in men.
In addition, I was newly married to a Yale graduate, liable for military service in the British army, and I had no job. I was grateful to be offered two very different jobs in Britain – one working on rocket fuel development and the other on the biochemistry of viruses at NIMR. I jumped at the NIMR job and was lucky enough to be given exemption from military service. I had a 3 year appointment as a member of the Chemistry Division, not the Virology Division, since Sir Christopher Andrewes, the head of Virology, was resolutely opposed to any scientist being a member of his division. For him, virology was a medical subject, accessible only to medical graduates, and was not taught for example, in any undergraduate science course in Britain at that time.
My first project was to determine the nucleic acid content of influenza virus, which was known to be an RNA virus, but how much RNA was uncertain, and there was some evidence that the amount of RNA depended on the way the virus was grown. Neither was it certain whether the virus also contained DNA — such was our ignorance. I spent the next eighteen months determining the RNA content of two forms of the purified virus (spheres and filaments) by the extremely laborious method of hydrolyzing the viral RNA to mononucleotides, separating them by ion-exchange chromatography, and using UV absorption to determine the relative amounts of each nucleotide. The work was published with Alick and the head of my Chemistry Division as co-authors; my first paper with Alick of many [1], though not on interferon.
Towards the end of this period, I started discussing with Alick what I should do next. I had in mind a very ambitious project that involved labelling purified influenza virus with radioactive phosphorus and following it through the infectious process. In retrospect, it would have been a disaster — the radioisotope would have gone everywhere and nothing interpretable would have resulted. He suggested, as an alternative, that I might like to help him “with something interesting that we are doing on interference.” “We” was Jean Lindenmann and himself, the time was a March, 1957, and interferon was only a few weeks old. Jean had, I believe, suggested the name interferon – Alick once complained to me that he thought that it was “time that biologists had a fundamental particle, for the physicists have so many; such as electron, neutron proton etc.”
However that did not stop Lord Hailsham, a senior lawyer, trained in the classics and the Chairman of the MRC at that time, objecting that it was a nasty hybrid word with both Latin and Greek roots! By then, though, the name had stuck. Alick and Jean worked well together; they had adjacent rooms on the second floor of the Institute, and since Alick was also Director of the World Influenza Centre, he had a large laboratory that tended to be our communal workspace. Jean supplied an admirable foil for Alick’s mercurial, effervescent temperament, while I had a chemical training which came in useful.
Interferon had been discovered by a series of experiments planned to test quite another hypothesis. It was the early days of virology (the steam age, as Sir Christopher Andrewes would say, referring rather disparagingly to the dream age that would follow – molecular biology and all that which he did not believe in!), and no one really knew how animal viruses worked – indeed it was suggested that the viral coat was left outside the cell, as did bacteriophage. Alick and Jean were testing this by seeing whether any viral property – and they chose interference – was still associated with the outer coat membrane of the cell, and could be washed off. What they found was not the viral coat from outside the cell, but the interferon newly made inside the cell. It was their perception that there was something unusual going on, when the small effects that were observed could easily have been dismissed as experimental error, as well as the formulation of a testable hypothesis that was the real insight.
The system was very crude. The virus that was used to stimulate interferon production was heat-inactivated influenza virus, which could interfere but not multiply, and the cells used were pieces of chorioallantoic membrane, cut from a 10-day-old fertile hen’s egg. The virus preparations were not very potent and one of the improvements we made soon after I joined the collaboration was the use of ultraviolet-inactivated virus instead of heat-inactivated virus. Interferon was estimated by challenging the treated cells with infectious influenza virus and then measuring virus growth by hemaglutination titration.
Influenza has the capacity to bind to chick red cells, so called agglutination, and finding the dilution of virus that gave partial agglutination provided a simple, though insensitive, measure of the amount of virus present. It was necessary to test, in duplicate, at least three two-fold dilutions of the interferon sample to get a response onto the dose-response curve. The amount of virus produced in interferon-treated cells was measured by diluting the virus produced in serial two-fold steps in plastic plates, and then adding chicken red blood cells. The endpoint of the titration was the well with partial agglutination, and a reciprocal of the interferon dilution, the interferon titer.
The experiments took hours to titrate, involving little more than purely mechanical operations, and this left time to talk. Alick was the leader in conversation, and ideas for new experiments, political discussion, or identification of snatches of opera that he would sing made the time pass quickly. Alick, too, was adept at determining where the endpoint of the titration was, and with the aid of a hand lens, could do it long before the rest of us, so he had often planned the next experiment before the red cells were really settled. Occasionally, in his impatience to start the next experiment, he mistook the endpoint and then the experiment was abandoned before it had even been started! It was immensely stimulating, and very different from the chemistry I had being doing, for that was a mature discipline, and this was so new!
By March, 1957 Alick and Jean had established the basic phenomenon and, together with the electron microscopist Robin Valentine, had looked hard for virus particles in the interferon preparations, since it was quite possible the interference detected in the fluids was due to residual virus particles. Two papers were written, and were sent to the most prestigious journal of the time, the Proceedings of the Royal Society, but there was still much to do.
The first of these two papers [2] described the production of interferon by treatment of chick chorioallantoic membranes in vitro with heat-inactivated influenza virus, and went on to show that interferon could be distinguished from heat-activated influenza virus by several properties; interferon was non-hemagglutininating, its activity was not neutralized by viral antiserum, and it was not sedimented by high speed centrifugation. However it was not possible to decide whether interferon was a cellular product formed in response to virus infection, or a part of the heated virus itself, or whether possibly it represented an abortive attempt at virus multiplication. The second paper [3], which was quite short, described attempts at visualization of interferon.
It was quite easy to plan and carry out the experiments that characterized the system further, and these were published in a series of papers in the British Journal of Experimental Pathology. I still have my laboratory notebooks from those early years, and my first experiment, dated March 4th, 1957 was headed “Dialysis of interferon” – we did not even know whether interferon would pass through a dialysis membrane or not! A second experiment, started on the same day, was to test whether interferon activity was destroyed by shaking a crude preparation with ether. It was, and it was another hint that interferon was a macromolecule. A series of experiments to characterize the stability of interferon at different pH’s followed and then several experiments to see whether interferon did really behave like a macromolecule, either a polysaccharide or more likely, a protein.
I found that it was precipitated with ammonium sulphate, (experiments carried out in early May, 1957), and that it was degraded by treatment with the proteolytic enzyme trypsin, and also that it was inactivated by shaking with butanol but not inactivated with periodate – suggesting it was not a polysaccharide. The stability of interferon at pH2 gave us an opportunity to test whether it was destroyed by the proteolytic enzyme pepsin, and it was, confirming that interferon was a protein, and if it was a protein, then presumably it could be purified, possibly relatively easily. The first of these conclusions was true, but the second took a long time and was much more difficult.
The first paper of the series in the British Journal of Experimental Pathology with the title: “Studies on the production, mode of action and properties of interferon” [4], was the only paper for which Jean and I were coauthors with Alick, before Jean went back to Switzerland in September 1957. Alick wrote papers very quickly; he would take the laboratory notebooks home and produce a first draft by the next morning, and so we were able to submit this paper as early as July 23rd, 1957. In brief, it described a system that we used to make interferon for the next few years, some experiments showing a need for cell metabolism before interferon could be effective and others showing the lack of specificity of interferon’s action.
It also described the well-known pH2 stability, precipitation by ammonium sulphate, and inactivation by trypsin. Thus this early paper established a substantial number of the basic parameters of the interferon system. We had, of course, no idea as to how complex the system was going to be – Alick and Jean’s first paper had been called “The interferon”, as if it was a single substance, and it was not until David Tyrrell later showed that interferon was often species specific that we realized that there was more than one interferon.
The next paper, rather frugally titled “Further studies on interferon” [5], was submitted on November 7th, 1957, and described the use of ultraviolet inactivated virus as a much more efficient producer of interferon and showed that the time of a irradiation was very important in determining the yield, small amounts of irradiation producing high yields, whereas longer periods of irradiation led to a complete loss of effectiveness. These experiments are now most readily interpreted as a measure of the capacity of the virus to form double-stranded RNA which was, in turn, the actual inducer.
The final paper in that early series was modestly called “Mode of action of interferon” [6]; it seems incredible, looking back, that we could have thought that the problem was that simply solved. This short, rather complicated paper, showed that per-treatment of cells with interferon, followed by the induction by inactivated virus led to an increased yield of interferon, a phenomenon called ‘priming’. This effect has now been explained by the induction of otherwise rate-limiting transcription factors required to produce interferon messenger RNA.
However we knew nothing at that time about transcription factors of course, and at the time we advanced a rather complicated, though ingenious, interpretation of what we had observed. Rereading the paper after all these years, it strikes me that the conclusion of that paper is remarkably dense and strikingly void of any molecular interpretation. It is of course more a comment on how descriptive our understanding of cellular processes was at that time.
In the event, these results were pushed to one side, when a simple and elegant experiment by Joyce Taylor in 1964 showed that interferon production was inhibited by treatment of virus-infected cells with actinomycin, and since it was known that actinomycin blocked DNA-directed-RNA synthesis, and since the interferon was induced by infection with an actinomycin-resistant virus, it was clear that cellular DNA must be involved. Though that explained the cell specificity of interferon very neatly and provided no insight as to the actual process, it did provide the essential molecular framework for much of the work that followed in the early sixties.
One of the most striking characteristics of the interference phenomenon, which we now believed was mediated by interferon, was that one virus can interfere with the growth of a number of unrelated viruses. It was therefore important to see how broad the antiviral effect was and the next paper in the series [7], published in October 1958, showed that the chick chorion of the 12-day old fertile hen’s eggs could be used to measure the protective effect of interferon against vaccinia virus.
A similar protective effect was found against two other poxviruses, cow-pox and ectromelia, although herpes simplex appeared to be more resistant. This was the first demonstration that interferon was active in vivo, although only in the fertile hen’s egg, not in an animal. More importantly, it really did look as if interferon had a wide specificity and this raised the important practical question of whether interferon could be developed as an antiviral antibiotic. The last paper in the series reported on “Some factors affecting the production of interferon” [8], showing a general correspondence between the capacity of influenza virus to produce interference and interferon.
By then interest in interferon was growing and already the focus was shifting to the possible utilisation of interferon. Virus infections were very important medically, there were no antiviral drugs and vaccine development was in its infancy. Alick Isaacs and I wrote a general article titled “Interferon: A possible check to Virus Infections” [9] which was published in the British weekly science journal The New Scientist in June, 1958. Interferon even made the Flash Gordon cartoon! We were also honored by an invitation to present our results at a Conversazione (a reception with food, wine and scientific exhibits) for the Fellows of the Royal Society (Alick was elected a Fellow in 1966) in May 1958, and I have a copy of our abstract which started:
“So far no antibiotics active against viruses have been discovered. To a large extent this is because viruses are extremely small parasites which are obliged to live inside cells, and it has not been possible to find a substance which would stop viruses from growing without at the same time harming the host cells. Interferon is the name which has been given to a new substance which prevents the growth of a number of viruses without apparently causing any gross damage to the cells. Interferon does not kill the viruses, but stops them from multiplying. This demonstration shows different aspects of the study of interferon….”
We had a rather simple series of posters and demonstrations which showed this distinguished body of senior scientists, all Fellows of the Royal Society or their guests, some of the early results and its promise. We were all dressed up, quite appropriately, in dinner jackets, and I remember that we were asked to present our demonstration a second time to an event to which only the really ‘great and good’ were invited. For this event we had to wear white tie and tails, which I didn”t possess and had to hire. I vividly remember dressing up in our very modest little North London flat, and sitting down with my wife to eat in my splendor, and she complimented me by putting on an evening dress, as we sat at the kitchen table, before going off to the great event. It was a heady time; I was only 28.
However, some problems were surfacing. The first indication was a puzzling positive result – we could get protection against the growth of vaccinia virus in the rabbit skin using chick interferon. That did not cause us any concern at the time because it had still not been shown, as it was so clearly later by David Tyrrell, that interferon was species specific and that chick interferon was not active in rabbit cells. But as soon as it was, we had to ask ourselves why was our preparation of chick interferon preventing the growth of vaccinia in the rabbit skin? It struck us that this might be due to traces of ultraviolet inactivated virus coming through from the cells in which the interferon had been prepared, contaminating the interferon and adding to the interference affect.
If that was so, how many of the other results were due to traces of UV inactivated virus as a contaminant? This troubled us greatly, and it coincided with criticism of the interpretation of our results in the US, where interferon was being called “misinterpreton” and several eminent US virologists were dismissing the effects as due to traces of virus. Alick was very depressed by this reaction, and it was the first sign of a series of depressive setbacks which dogged him over the next few years. He was off work for a month or two and I spent that time repeating all the initial experiments with interferon which had been treated at pH2 in order to destroy any UV inactivated virus, so as to be quite sure that the effects we had been observing, and publishing, were due to interferon and not to traces of contaminating virus. To our relief, all the early experiments held up, and it was not necessary to publish any retraction or corrections.
Two lines of inquiry dominated our time for the next few years. The first was to see whether interferon could really be developed as an effective antiviral agent in the UK. In the late fifties the outcome of the penicillin story still grated in Britain; the perception was that a British discovery had been “handed over” to the Americans during the war, they had then developed an industrial production process which had been patented, and we were now paying royalties on that process in order to obtain the drug we had discovered.
So the MRC was under considerable political pressure to determine whether and how interferon could be developed as an effective antiviral agent in the UK. The actual discovery of interferon was patented, although the legal process was so cumbersome, and was so delayed as result of challenge from the United States, that the patent did not come into effect for some years afterwards. However that patent did provide some royalty income at a time when interferon was being developed on a large scale, so it was not wasted.
More to the point, a novel groundbreaking collaboration was built between the MRC and three major pharmaceutical companies working in the UK: Glaxo Laboratories, ICI Pharmaceuticals and Burroughs Wellcome, later to become the Wellcome Foundation. This was set up about 1958 and worked until the mid-sixties with the specific aim of making enough interferon to do an effective clinical trial. The collaboration brought new skills and new people into the field: Karl Fantes from Glaxo, and Norman Finter from Burroughs Wellcome were outstandingly valuable additions, and many other resources became available. For example, the standardization of the interferon unit, the development of better methods for large scale production, and experienced development management skills from the pharmaceutical industry all brought benefits.
I was a member of that Committee throughout its life, and Alick was chairman. He was not a good chairman; my experience is that research academics rarely have the necessary skills to steer a mixed academic/industrial project forward, and in retrospect a professional, experienced manager from one of the pharmaceutical companies should have chaired the whole process, but we were feeling our way at that time towards effective research/industry collaborations and what is obvious now was not obvious then.
Partly because of Alick’s style and partly because of his series of illnesses, the collaboration had its up’s and down’s, and on a number of occasions went off down blind alleys, but it did achieve its initial objective of a trial against a vaccinia virus challenge in the upper arm of unvaccinated volunteers at the Common Cold Research Unit at Salisbury in the spring and summer of 1962. So the outcome was two edged: on the one hand, the collaboration had shown that interferon could be used in humans against a virus challenge, but on the other hand, it was not practical to prepare either enough interferon, or to deliver it early enough to be a either a useful prophylactic or a therapeutic.
So other systems to determine the parameters had to be explored, and herpes infection of the rabbit eye was one system which did give a useful clinical outcome. But the whole clinical development of interferon was put on hold then for some years, partly because of our inability to make enough interferon – a problem not solved until the development of large scale production in human cells by Kari Cantell in Helsinki, using human leukocytes, and by Norman Finter in the UK, using human lymphoblastoid cells, and finally by the production of interferon by gene cloning in the early eighties. The other new driver which emerged in the seventies was the claim that interferon could be used effectively against human cancer, but that is quite another story.
The other line, which was my responsibility, and filled my time until the early sixties when I began to work on other aspects of interferon production, was its purification. Early experiments had shown to our satisfaction that it was a protein: suitable purification procedures were then being rapidly developed, and there was expertise available in NIMR, especially in the group working around Rodney Porter, who was awarded the Nobel Prize for elucidation of the structure of antibodies from work he did at this time.
So all looked good. The object was twofold: to prepare material that was could be used in clinical trials, and to establish exactly what sort of physico-chemical entity interferon was. In the event, this took years and the story became increasingly complex as it emerged that there was not just one human interferon but many: α, β and γ, and also multiple varieties of interferon α and all this was unknown when I started serious work on interferon purification in the summer of 1958. Nor had we any idea how high the specific activity of interferon would turn out to be: our best preparations had about 1000 units per ml and with a specific activity of about 109 units per mg; we had only about ten micrograms of the interferon in our 10 litres of starting material.
So although we scaled the process up, ultimately working with ten litre batches, the amount of material we were trying to purify was very small, and because interferon was readily absorbed on to surfaces or became attached to other proteins present in the crude preparation, and also because it emerged that column purification procedures would only work effectively with high loads of interferon protein, it is clear that in retrospect that the desired outcome, that of making and characterizing pure interferon in a year or two was a hopeless task.
But we pressed on and scaled up in using larger and larger bottles and more and more eggs but still using the initial process of treating chorioallantoic membranes with ultraviolet inactivated virus. This was before the use of tissue culture systems which were then being developed on an industrial scale for polio vaccine production, and it was long before the production on the multi-thousand litre scale, developed by the Wellcome Foundation for the production of human interferon. We did all our own assays, and Friday, Saturday morning and all day Monday was taken up with the assays from experiments that had been run on Tuesday and Wednesday.
It was a difficult task and looking back, it was impossible, but of course that is hindsight with its 20:20 vision. So I filled twelve laboratory note books with experiments aimed at developing a multi-stage purification process for chick interferon. I established a partial process by the spring of 1960, just before I left NIMR at the end of March, 1960 to take up a university lectureship. It was, as the Duke of Wellington said of the battle of Waterloo “the nearest run thing you ever saw in your life” since my contract expired in the summer of 1960, and with a wife and two children, I had to get a job.
My original three year contract had been extended for another two, mainly as the Director rather tactlessly explained, to keep the lab going while Alick was unable to work. So I had no long-term future there and I looked for a job where I could continue working on interferon and the biochemistry of viruses, and I needed another virologist in the same university to help get me going. In 1960 that restricted my choice to just one or two universities, and there were only a few jobs going every year in biochemistry anyway. So I was fortunate to get a lectureship in biochemistry at the University of Aberdeen to start in April 1960, and that gave me a very firm deadline for the completion of the purification work.
I vividly remember taking the last ten litre batch through the purification process desperately hoping that nothing would go wrong. And in those days it often did. The fraction collectors, which were essential for collecting the eluate from the ion exchange columns we were using for the multistage process, were made in the Institute workshop and were unreliable because they worked on a siphon system, the filling of the siphon triggering the move to the next test tube. But as soon as protein started to be eluted from the column, the surface tension changed, the siphon started siphoning continuously and the crucial eluate went all over the cold room floor. So I used to work late at night watching over the fraction cutter, and get into the laboratory as early as I could, often about six am when I was doing a big run.
We did manage to get enough material through the multistage procedure from the final ten litre batch to give us enough biologically active product to characterize by starch gel electrophoresis – polyacrylamide gels had not been invented – and to do an analysis in an analytical ultracentrifuge. The material was homogeneous on both counts, had a molecular weight of about 63,000 and we really thought we had a homogeneous product [10]. At this point the removal van had taken away our furniture to Scotland, and we were living in our small flat with two children on a day to day basis desperately trying to finish before I had to drive to Aberdeen with the family. So we thought we had made it, but the product turned out not to be pure interferon but chicken albumin, to which some interferon was hydrophobically bound.
There was indeed one protein, and it was associated with biological activity, but it was not pure interferon, and it took some years before other workers in United States and Britain completed the task. After one modest attempt to continue purification work in Aberdeen, I decided this was a project that was impossible to continue in an academic setting, and switched to a study of the mechanism of interferon production. However, my paper had unequivocally shown that the biological activity of interferon was associated with a purifiable protein, and that interferon was not just a figment of the imagination which, like the Cheshire cat, faded way as soon as it was inspected.
But back to 1957. It was a very special summer; for it is not often that at 27, one can publish effectively every experiment, and that of course helped me when I did come to leave NIMR. Over the three years 1957 to 1960, interferon had been firmly established as the mediator of virus interference, as a protein which was purifiable, and as an important new lead in dealing with virus infections. It was also a marvelous learning time for me personally, trained as a chemist, working with outstanding biologists for the first time. Interferon was new, exciting and had clear medical applications.
However, best of all was the company, and I shall always remember Alick, Jean, and I doing hemagglutination titrations in room 215. It was also the summer our first child was born – I remember that we called her the “interfering particle” because of lost sleep. The years slip by, and now that she is over 50, I’m reminded of the discovery that Jean and Alick made 50 years ago.
Author(s) Affiliation
D Burke – 12, Cringleford Chase, Norwich, Norfolk NR4 7RS, UK email: dcb27@cam.ac.uk
References
-
- Burke DC, Isaacs A, Walker J. The nucleic acid content of influenza virus. Biochim Biophys Acta, 1957, 26: 576-584.
- Isaacs A. Lindenmann J. Virus interference. I. The interferon. Proc Roy Soc, Ser.B, 1957, 147: 258-267.
- Isaacs A. Lindenmann J. Valentine RC. Virus interference. II. Some properties of interferon. Proc Roy Soc, Ser. B, 1957, 147: 268-273.
- Lindenmann J, Burke DC, Isaacs A. Studies on the production, mode of action and properties of interferon. Br J Exp Pathol, 1957, 38: 551-562.
- Burke DC. Isaacs A. Further studies on interferon. Br J Exp Pathol, 1958, 39: 78-84.
- Isaacs A. Burke DC. Mode of action of interferon. Nature 1958, 182:1073 -1074.
- Isaacs A, Burke DC, Fadeeva L. Effect of interferon on the growth of viruses on the chick chorion. Br J Exp Pathol, 1958, 39: 447-451.
- Burke DC, Isaacs A. Some factors affecting the production of interferon. Br J Exp Pathol, 1958, 39: 452-458.
- Isaacs A, Burke DC. Interferon: a possible check to virus infections. The New Scientist, June 5th, 1958. Burke DC. The purification of interferon. Biochem J, 1961, 78: 556-564
Related stories you may like:
Meningeal Interferon-γ Affecting Neuronal Connectivity and Social Behavior
Interferon-γ Levels and Chronic Fatigue Syndrome
Hyperactive Interferon System Choroid Plexus Linked To Aging and Cognitive Dysfunction
Anxiety Linked to High IFN-γ and TNF-α
