Interview with the founding director

Prof. Dr. Manfred F. Rajewsky

For many years I was most interested in the process of cancer development.

Later, I began to think more about ways to improve cancer therapy. There had been cancer research since the beginning of the century. As with infectious diseases, people had always hoped to find a kind of philosopher’s stone, a mechanism by which cancer always develops in the same way. However, no such mechanism had been found. So when I started, there was a certain resignation, which was also communicated to young people like me.

I had done my doctorate at a laboratory whose name stood for having investigated the development of cancer after exposure to chemical compounds very precisely, especially with regard to the relationship between dose and effect. Hundreds of very different compounds (carcinogens) turned out to be carcinogenic. I found it hard to believe that all these chemically very different substances should always transform normal cells into malignant cells in the same way.

However, I was much more interested in a phenomenon that was rarely discussed at the time, namely the fact that many chemical carcinogens cause cancer to develop predominantly in certain tissues or organs. Certain carcinogenic substances and certain tumors belong in the same drawer, so to speak: One substance produces ear canal tumors, the other liver carcinomas. This phenomenon is called organotropy of the carcinogenic effect.

At the time, I thought that we should pay much more attention to this organotropy because it contains an element of specificity with regard to the mechanisms of carcinogenesis. There obviously had to be important determinants in the process of carcinogenesis that lay not only in the reactivity of the respective carcinogen itself, but also in different properties of the affected cells. To prove this, one would need a substance that demonstrably reacts in the same way with all cells in the organism. If the tumors then still only developed from certain cell types, but were not statistically distributed throughout the organism, then one could be sure that different mechanisms are set in motion on contact with carcinogenic substances depending on the cell type, because the cell type is the only variable in this model experiment.

So while we were still in Tübingen, we started looking for an organism and a suitable carcinogen as a model for these analyses. We finally selected a substance, N-ethyl-N-nitrosourea (EtNH), one of the many carcinogenic N-nitroso compounds, which is highly carcinogenic after a single dose and predominantly causes malignant tumors of the central and peripheral nervous system in rats. We systematically measured the reaction products of EtNH with cellular macromolecules, especially with DNA. This was still very time-consuming at the time. The carcinogen had to be radioactively labeled in order to be able to find the reactive group at the site of its binding. Such labeling was difficult and cost about DM 10,000 each time, so we had to rely on outside help. The Radiochemical Laboratory of Hoechst AG has always generously supported us, also financially. We used radiochromatography to measure the reaction products with the DNA until we knew in which tissues which products were formed in which quantities, and that these values were always almost the same. This was our first prerequisite.

In our model, however, the carcinogen had to have another property, which fortunately it did: a single dose (pulse) had to be sufficient to induce a large number of tumors. Because if you want to analyze a multi-stage process such as carcinogenesis, it is very disturbing if you need multiple or continuous exposure to the carcinogen, because then the initial stage of the process would be triggered again with each single dose, with the consequence that it would be difficult to distinguish initial from secondary mechanisms.

Another important property of this carcinogen was added. In addition to the hoped-for organotropy, the carcinogenic effect was also strongly dependent on the timing of the carcinogen pulse during the development of the nervous system. The probability of tumor development was thus also dependent on the developmental stage of the target cell system.

Prof. Dr. Manfred F. Rajewsky

Yes, certain properties that different cell types in the organism have depending on their stage of development and differentiation at the time of carcinogen exposure and which determine their risk of being transformed into cancer cells. Such studies must be carried out on well-defined model systems, since under natural, i.e. environmental conditions, carcinogen mixtures are always present whose mechanisms of action are difficult to analyze in detail. This is another reason why cancer research is such difficult terrain.

Prof. Dr. Manfred F. Rajewsky

At this time, the central role of DNA was already clear. The analysis of carcinogen-induced DNA damage and its possible consequences for the cells, e.g. mutations in critical genes, began to become increasingly important.

We knew from experience with our experimental model system that certain structurally well-defined DNA damage occurs in dividing cells, which is highly likely to trigger mutations in the subsequent rounds of DNA replication. It was clear that the mutation frequency would be reduced if the DNA damage was repaired before replication. Was the organotropy of the carcinogenic effect of EtNH related to a different DNA repair capacity of the cells hit? Were cells with a particularly low repair capacity characterized by a high transformation risk, i.e. a high probability of transformation into a tumor cell? We then compared different cell types in terms of their ability to remove damage from their DNA.

Today we know that a large number of cellular proteins are involved in different DNA repair processes. At that time, nobody knew of a single DNA repair protein. So we looked at the end result of the presumed DNA repair. The repair capacity of the cells was measured by how much of the initially induced DNA damage disappeared from the DNA as a function of time. We quickly discovered that – in contrast to other DNA reaction products – a very specific DNA modification (O6-alkylguanine) is hardly repaired by the cells of the nervous system, from which EtNH-induced tumors arise. The cells of other tissues were repair-active, albeit to varying degrees. O6-alkylguanine is a particularly mutagenic DNA modification. Thus, we were able to show for the first time that DNA damage induced by chemical carcinogens is repaired by different cell types with different efficiencies and that this different ability to repair DNA appears to have something to do with the transformation risk of the cells.

Prof. Dr. Manfred F. Rajewsky

Of course, we had been following these studies with interest. From a completely different perspective, there was a strong indication of a connection between the inability of cells to repair certain mutagenic DNA damage and subsequent tumor development.

Prof. Dr. Manfred F. Rajewsky

For the start in Essen, we had set ourselves the goal of first of all significantly refining DNA analysis. If we wanted to make progress, we had to significantly increase the sensitivity of detection for specific DNA changes. Quantitative measurements of even very small amounts of a defined DNA change in small cell samples, single cells, even in individual genes, had to become possible. Today, in addition to highly sensitive DNA repair analysis, we have a molecular epidemiology for specific DNA damage in humans caused by exogenous agents. It is of considerable importance for cancer risk assessment as well as for the molecular dosimetry of DNA-reactive chemotherapeutic agents in cancer patients and the pre-therapeutic detection of DNA repair-related therapy resistance. At that time, it seemed hopeless to pursue these areas without improving the methods for measuring specific DNA changes.

Prof. Dr. Manfred F. Rajewsky

They were just starting to do that. But before that, we had thought very fundamentally about whether it might be possible to use antibodies for the specific detection of carcinogen-induced DNA changes.

Prof. Dr. Manfred F. Rajewsky

Yes, an antibody is a molecule that recognizes molecular structures of foreign origin very precisely, i.e. it binds to these structures. We had to obtain antibodies that recognize and bind to bases in DNA that have been structurally modified by carcinogens. The problem was that in the case of our model carcinogen it was an extremely small change in the DNA: just an additional methyl or ethyl group attached. But of course we tried it anyway, around the time when the first monoclonal antibodies were produced by Köhler and Milstein. In 1975, they discovered that a B lymphocyte that secretes a certain antibody can be multiplied at will by fusing with a tumor cell and the daughter cells then continue to produce the same monoclonal antibody. The classic antisera, on the other hand, are much less precise; they are mixtures of many different antibodies. In contrast, monoclonal antibodies are ultra-pure recognition tools.

We then very soon started to produce monoclonal antibodies and immediately hit the jackpot. One of our first attempts to produce a monoclonal antibody for the specific recognition of a DNA alkylation product (O6-alkylguanine) was a complete success. Our antibody actually had the properties we had hoped for: extremely high specificity and binding affinity to the antigen. We had thus opened up a new analytical world. With the help of monoclonal antibodies and the immunoanalytical methods we developed, we were able to make measurements that were orders of magnitude more sensitive.

Prof. Dr. Manfred F. Rajewsky

We published the first report in 1980. We had been working on these results with the help of many capable employees since 1976, after we had started to set up our Essen laboratories in 1975. In 1985, we set up our central laboratory unit for monoclonal antibodies because it quickly became clear that we needed to produce more and more monoclonal antibodies against various carcinogen-induced DNA modifications. In the meantime, we have produced many hundreds of monoclonal antibodies, so that today our institute has the largest arsenal of monoclonal antibodies for the identification of carcinogen-induced DNA modifications in the world. Many important DNA repair analyses on single cells, more recently also at the level of specific genes, have subsequently been carried out. Numerous domestic and foreign scientists were involved in this work.

With the start of your work, the West German Tumor Center Essen gained a strong foothold in basic research. How did the cooperation between the IFZ and the clinical departments develop?

First of all, we had to make an effort to establish good basic research in an existing center of clinical oncology. A so-called Comprehensive Cancer Center, as the Americans call it, can only bear this title if it has both clinics and basic research in equal measure. This did not exist in Essen at the time. It was clear to us that we were starting here with a big mortgage: We were required to provide the necessary impetus in basic research.

Prof. Dr. Manfred F. Rajewsky

No, by no means just us. Professor Heinrich Schulte-Holthausen was appointed head of a second new institute, the Institute of Molecular Biology (tumor research), almost at the same time as me. We were in the same building, on the same floor, and each of the two institutes had a corridor at its disposal. We had no lecture hall, no seminar room and laboratory rooms, which we first had to set up.

The IFZ then owed a great deal of progress to the Alfried Krupp von Bohlen und Halbach Foundation, which made it possible to set up a seminar and course room with a reference library in 1981 thanks to a generous donation. Looking back, this donation was decisive for our development at that time: without these facilities, the IFZ would not have become the point of communication for research at Essen University Hospital that it is today. We were also able to expand our laboratory space with the support of the Alfried Krupp von Bohlen und Halbach Foundation.