Genetics, Vol. 167, 555-558, June 2004, Copyright © 2004
doi:10.1534/genetics.104.030320

Leo Szilard

A Personal Remembrance

Werner Maas¹

Department of Microbiology, New York University School of Medicine, New York, New York 10016-6402

¹ Address for correspondence: Department of Microbiology, New York University School of Medicine, 550 First Ave., New York, NY 10016-6402.
E-mail: maaswo1{at}endeavor.med.nyu.edu

SINCE retiring as a full-time faculty member about 10 years ago, I have occupied myself with writing chronicles of past experiences. The first was an account of the advances made in the elucidation of gene expression and its regulation during the twentieth century (MAAS 2001). The second was an account of the history of a Jewish family in the Palatinate region of Germany, based on a memoir written by my father after World War I (MAAS 2003). Now, in looking for another chronicle to write, I was very pleased to be asked to write a Perspectives essay about Leo Szilard. I could think of no better topic and I accepted enthusiastically.

I first met Szilard in 1951 and had many contacts with him until the time of his death in 1964. During this period I developed an unreserved admiration for him as a scientist and as a human being. I consider it a privilege that I now have an opportunity to describe this unusual man as I remember him.

Szilard started a new career in biology in 1947 after having made significant contributions to physics and having been instrumental in the development of the atomic bomb. At that time he asked Aaron Novick, a young physical organic chemist, to join him in his new venture. Novick accepted enthusiastically, and this was the beginning of an association that lasted over many years. They began their entry into biology by taking the phage course at Cold Spring Harbor in 1947, which at that time was taught by Max Delbrück (see SUSMAN 1995).

Novick and Szilard set up their own laboratory at the University of Chicago in 1948, where Szilard had been appointed a professor of biology and sociology. They became members of the new Institute of Radiobiology and Biophysics, which was located in a former synagogue of a Jewish orphanage. Their laboratory was in the basement of this building. Their plan was to study mutation in bacteria, and they looked for methods to measure mutation rates.

At Cold Spring Harbor they had encountered Jacques Monod, who had told them about his diauxie experiments in which, in the presence of the two carbon sources glucose and lactose, glucose is used preferentially. Szilard was intrigued by this finding, and to analyze it he and Novick designed an apparatus for the continuous culture of bacteria, which they called a chemostat (NOVICK and SZILARD 1950). Subsequently, this device turned out to be very useful for the study of mutation rates, as well as for physiological studies on bacteria. Aaron Novick has given a vivid account of his association with Szilard during this period (NOVICK 1966).

In 1951 I was working in the laboratory of Bernard Davis at Cornell Medical College in New York City. We were studying various biosynthetic pathways of Escherichia coli with the use of mutants isolated in our laboratory. Szilard was a frequent visitor and liked to discuss our work with us. My own problem concerned the manner in which genes control the production of enzymes, as postulated in Beadle's one gene–one enzyme hypothesis. Do they determine the structure of enzymes or merely the rate at which the enzymes are produced? To study this question, I used a temperature-sensitive mutant requiring pantothenic acid for growth in which the enzyme affected by the mutation could be measured in extracts. I found that the extracted enzyme was much more heat labile than the corresponding enzyme from the wild-type strain. It seemed likely that the enzyme molecule itself was affected by the mutation, but there was the possibility that the observed difference was due to an extrinsic factor produced in the mutant. I posed this question to Szilard, and he immediately suggested that I could get an answer by studying heat inactivation in mixtures of mutant and wild-type enzymes. I carried out the experiments Szilard suggested, and the results showed clearly that the difference in thermal stability was due to a difference in the enzyme molecules themselves.

I am ashamed to say that I did not mention Szilard's crucial suggestion in the acknowledgment at the end of my published paper (MAAS and DAVIS 1952). I can say only that, as far as I know, there are other people who received similar crucial suggestions or advice from Szilard without acknowledging his contribution. I think that Szilard was not bothered by these omissions, because his passion was to solve scientific questions, and he was not concerned about receiving personal credit for his contributions.

Szilard spent a great deal of his time visiting other scientists in the quest for obtaining information about questions that interested him. He was very purposeful during his visits. As an example, during a party in Bernard Davis' apartment in New York City, in which Monod, François Jacob, Novick, and other researchers were present, Szilard suddenly appeared. However, he did not join the group in the living room, but instead took over a bedroom and invited each guest, in turn, in for a private chat, quizzing them on their latest work and findings, suggesting new experiments and novel interpretations, and reporting how these might relate to the work of others.

Szilard was a very private man. He never talked about personal problems and even did not talk about scientific problems, unless he could formulate them in precise questions. I experienced these aspects of his personality when I visited him in the mid-1950s in Denver, where he was spending the summer carrying out experiments at the University of Colorado Medical School. He asked me to join him for lunch and to first meet him at his laboratory at the Medical School. When I arrived, he was working with a column, apparently carrying out a separation. When I asked him what the experiment was, he would not tell me. I found out only later that he was trying to separate male-producing sperm from female-producing sperm to be able to control the sex ratio. One of his vital concerns was human reproduction and its consequences for the welfare of society.

In retrospect it was perhaps just as well that he did not tell me about the nature of his experiment. In a recent article in The New York Times Magazine (November 1, 2003) there was a story about a woman who, after having had three boys, wanted very much to have a girl. She had been told that a method for being fertilized with female-producing sperm from her husband was available. The couple decided to undergo this procedure and the woman became pregnant, but the pregnancy resulted in the birth of male twins. After 50 years, Szilard's reticence was vindicated.

His sense of privacy was also revealed to me in Denver. I had been wondering why Szilard spent his summers in Denver in the first place. I learned only much later that he had a wife who taught public health at the University of Colorado Medical School. Her maiden name was Trude Weiss and he had married her in 1951. Nobody I knew, including Szilard himself, ever mentioned this relationship to me.

Our subsequent lunch engagement had an amusing aspect. We ate at a rooftop restaurant of what seemed to be Denver's only skyscraper. When we sat down I noticed that all the other tables in the restaurant were occupied by women. Szilard seemed not to notice this and we carried on our discussion. After a while some well-dressed women were walking slowly among the tables and I then realized that we were at a fashion show. One of these ladies came to our table and after looking at us said, "I guess you guys are not interested." The remark did not bother Szilard and we continued our discussion.

My next encounter with Szilard occurred in 1956 at New York University School of Medicine where I was studying the control of enzyme synthesis in the pathway of arginine biosynthesis in E. coli. I had discovered that the addition of arginine to a growing culture inhibited the formation of the enzyme ornithine transcarbamylase (OTCase) that catalyzes the conversion of ornithine to citrulline, an intermediary step in arginine biosynthesis. A former colleague from the Davis Laboratory, Henry Vogel, had found inhibition by arginine of the formation of another enzyme of the arginine pathway and had coined the term "enzyme repression" for this feedback control.

While studying the rate (kinetics) of OTCase formation following the removal of arginine from a growing culture of E. coli, I found an unexpected result. At that time Monod was studying the kinetics of enzyme formation for the formation of the inducible enzyme ß-galactosidase after addition of lactose. He had found that the enzyme is produced at a constant rate for several hours. In contrast, I found that OTCase was produced at a fast rate for only 20 min after the removal of arginine, after which there was a sharp drop in the rate of enzyme synthesis. To explain this unexpected result, I postulated that during the initial 20 min arginine would be made rapidly in the cells and the accumulated arginine would inhibit further OTCase synthesis.

At that time I was joined by a visiting investigator from Italy, Luigi Gorini. To test the idea that the decrease of OTCase synthesis was due to accumulated arginine, we had to find conditions of growth under which arginine could not accumulate. We discussed this problem with Szilard and Novick, who happened to be visiting our laboratory, and they suggested that such conditions could be obtained by growing an arginine-requiring mutant in a chemostat, with arginine being the rate-limiting growth factor. They had carried out chemostat experiments on the effect of added tryptophan on tryptophan synthesis and were in a good position to give us precise directions. When we carried out the experiment as they had suggested, we found that OTCase was produced at a high and constant rate, as expected from "Monod kinetics." We concluded from these experiments that the bacteria have a very high capacity for making OTCase, but that the formation of the enzyme is slowed down by internally produced arginine to a low level that is still sufficient for normal growth. Later this kind of regulation was demonstrated in many metabolic pathways.

My next encounter with Szilard occurred in 1957 when I attended the Federation Meetings in Chicago and stayed at the Quadrangle Club where Szilard was also staying. I met him during breakfast and told him about a conclusion I had reached on the control of enzyme formation on the basis of my experiments. It contradicted a widely accepted hypothesis proposed by Monod.

Monod had proposed that for inducible enzymes, such as ß-galactosidase, the gene coding for this enzyme would make an inactive enzyme protein, which would then become activated by interaction with its substrate lactose. Monod was a strong believer in a unitary hypothesis for the control of enzyme formation. Thus, for repressible enzymes, such as OTCase, he would postulate that arginine prevented the interaction of inactive OTCase protein with ornithine, the substrate for OTCase.

In our chemostat experiments we used a mutant blocked in ornithine formation, and yet this strain produced OTCase at high levels in the absence of its substrate. The conclusion I presented to Szilard was that for a unitary hypothesis one had to assume that, in the case of inducible enzymes, there was a repressor present and that the inducer prevented the action of the repressor. Szilard was impressed with my hypothesis and said that it could explain recent results they had obtained in their own laboratory. He wanted to publish my hypothesis right away, but I refused, because I had no direct experimental evidence for it. He did publish an article on the subject subsequently in 1960 (SZILARD 1960).

Early in 1958 Szilard presented my alternative unitary hypothesis at a seminar at the Pasteur Institute in Paris. At that time Arthur Pardee, Jacob, and Monod had begun an experiment designed to distinguish between the two hypotheses. A lively discussion followed the seminar, with the thinking of Monod clearly favoring his internal inducer hypothesis. When the experiment was finished and the results were analyzed, it was clear that the alternative hypothesis presented by Szilard was correct for the induction of ß-galactosidase. Enzyme induction was shown to involve the removal of repression.

Later in 1958 Szilard spent some time at the National Institutes of Health in Bethesda, staying at the nearby Kenwood Country Club. I visited him there and we had an animated discussion about the latest developments in the regulation of enzyme synthesis while sitting in the relaxing atmosphere of the swimming pool.

After 1960 I had only a few contacts with Szilard. He was heavily engaged in political activities concerned mainly with nuclear arms control and peaceful coexistence. At one time he invited me and my wife to his debate with Edward Teller on NBC on the topic, "Is Disarmament Possible and Desirable?"

In 1964, I visited Szilard in La Jolla, where he had moved to work at the Salk Institute. For several years he had been instrumental in creating this institute. He was living there with his wife Trude at the beautiful La Valencia Hotel. He met me at the San Diego airport and asked me immediately what was new in my work. "Nothing," I said. He looked at me, disappointed. "You think I am lazy," I said. "That is right," was his reply.

Szilard died in his sleep in La Jolla on May 30, 1964. To me it was a great loss and I shall always miss him. He had an extraordinary gift for rational thinking, which he used tirelessly and altruistically not only for the solution of scientific problems, but also for generating ideas and schemes for improving the social and political conditions of humankind. A detailed account of his life can be found in the biography by William Lanouette, Genius in the Shadows (LANOUETTE 1992).