|Home | About | Journals | Submit | Contact Us | Français|
Stimulated by Francis Cricks “central dogma of molecular biology” open to criticism and rebuttal, five statements are formulated on current notions of tumor biology, inviting the criticism of the reader.
In 1957, one of us (GK) participated in a 3-d meeting at The Burn in Scotland. The topic was “development, differentiation and cancer.” The first day was devoted to a series of talks by embryologists. But each speaker was interrupted by Francis Crick in the beginning. He said: before you tell your story, explain why you work with embryos.
It was very embarrassing. Of course, embryologists work with embryos. Do they have to justify this?
Renato Dulbecco was in the chair. After Crick has repeated the same procedure for the sixth time he said “Francis, I am the chairman here, and I will not allow you to ask that question again. But before you shut up, please explain what you mean.”
Somebody says to you: I have discovered an enzyme that is a polysaccharide. Your immediate response is: I do not believe it. The fact that enzymes are proteins is therefore firmly established.
Another example in the days when it was still debated whether RNA viruses carry their hereditary information in nucleic acid or protein, one of the pioneers of the field, Fraenkel-Conrat was to give a seminar about Tobacco Mosaic Virus (TMV). Shortly before starting his lecture, he received a telegram from his laboratory: be very careful what you say. The latest experiments cannot exclude that TMV protein may be infectious.
Did Fraenkel-Conrat speak more cautiously? No. He immediately knew that this was a practical joke. It was already firmly established that the genetic information of RNA viruses is carried by their nucleic acid.
How many equally firm principles are there in biology?—Crick continued. Perhaps 10, not more. It is only worthwhile to work toward establishing firm principles. Therefore, you have to select your objects carefully. If you are interested in development and differentiation, take bacteria where billions can be induced to sporulate at the same time. Or take a single somatic cell from a carrot plant and allow it to generate a whole new plant. In systems like this you may find out something about the regulation of development. An embryo is a mess; you cannot learn anything from it.
I commented “Nothing is a model for anything else.” Crick answered: “I could not disagree more profoundly.”
This was in 1957. Crick was, of course completely wrong. It would not take long before transgenic and knock-out mice would provide precise experimental tools for the study of development.
Crick liked to throw out provocative statements. Fully aware of the distaste biologists would feel for the word “dogma,” he formulated “the central dogma of molecular biology": “DNA makes RNA makes protein and it can never go backward.” This dogma was challenged when reverse transcriptase (RT) was discovered in the RNA tumor viruses. In this case, the backward move was from RNA to DNA. But the central dogma was still valid for cells. Nevertheless, the dogma and the discussions it elicited led to better understanding of genetic regulation. The “reverse flow of information “and its tool the RT led to a search for it in human tumors, assuming that its presence would be a marker for RNA viruses that could contribute to tumorigenesis. We think of the 1970s that followed as the “panviral decade,” a time of wrong concepts and flawed experiments. For a long time, you could not speak of DNA tumor viruses or chemical or radiation carcinogenesis without someone asking whether latent RNA tumor viruses were not the real culprits.
We can illustrate this with a personal anecdote. In the early 1970s, we were at a meeting in Jerusalem, staying at the American Colony Hotel. Sol Spiegelman also stayed there. One evening, we had dinner in the beautiful garden restaurant. At one point, he suddenly looked at his watch and said that he must go to his room and call his lab in New York. He was gone for a full hour. He came back beaming: The cancer problem is solved.
Sol Spiegelman was one of the great pioneers of molecular biology. He invented molecular hybridization, the very basis of all nucleic acid work. His technique has later expanded in a major way toward its great offspring, PCR. But Sol was not satisfied. It was just a technique. He had much greater ambitions. He wanted to generate a living organism in vitro, “to play God” as he said. His other goal was to solve the cancer problem. Now he solved it. It was all due to hidden retroviruses.
What did his people in New York tell him? They claimed to have found RT indicating the presence of RNA tumor viruses in every cell.
It was a giant artifact. They did not measure RT at all. Nevertheless, it created an enormous splash. A high security lab was built at NIH to handle potentially infectious cancer tissue. It took about 10 y before the emperor was declared naked and the paper tiger went to the waste paper basket. Still and in spite of the many blind alleys and erratic sidelines, this exercise was not useless. After the dust has cleared, a more realistic picture emerged. The omnipresent cancer viruses were gone—We may ask in retrospect: how could one have postulated them in the first place? Where would their selective advantage have come from? An important step was the discovery of the cellular origin of the oncogenes and their identification as growth controlling genes. Out went most of the virologists and in came the cell biologists. The remaining virologists settled in their well justified niches. Three herpesviruses; EBV, HPV and the Kaposi sarcoma associated herpes virus, HHV8, could obviously contribute to tumorigenesis. There was only one RNA tumor virus in the etiologically important category, HTLV1. None of these viruses have evolved to cause tumors. Their contribution to tumorigenesis was a byproduct of the viral strategy, such as S-phase induction in the host cell to establish the ground for viral sustainability, or of immunosuppression (EBV and HHV8) or of cytogenetic changes (papilloma viruses and HTLV1). The cell biologists came to occupy a broad platform. Tumor-related cellular changes were found to affect a whole range of properties, summarized by Hanahan and Weinberg under the umbrella of “Hallmarks of cancer,” twice with interval of a decade.1,2
We also want to exemplify the merit of provocative challenge by another major figure, the Australian Sir Macfarlane Burnet. Giving a talk in London, the chairman introduced him with the following words: here is Sir Mac again. Perhaps he will show us an experiment he has done on five eggs during 5 d, on the basis of which he has formulated a theory and wrote it up as a book. But now it will take 5,000 virologists, 5 y and 500,000 eggs to disprove the theory completely. By that time, however, a new science will have grown up.
Inspired by the examples of Crick and of Burnet, but in all modesty I shall attempt to formulate some statements on tumor biology. I invite the reader to challenge them, or to comment.
Statement 1. Illegitimately activated or constitutively expressed oncogenes may favor malignant transformation provided the cells are in a certain window of differentiation. If expressed in the proper window, they act by blocking the progress of the cell toward terminal differentiation or toward apoptosis, depending on the normal program of the cell. Deblocking may have a therapeutic effect. This argument is based on the early work of Holtzer and Boettiger.3-5 They showed that even temporary downregulation of the temperature sensitive Rous virus (RSV) oncogene v-src in RSV-driven cells, allows the cell to take an irreversible step toward differentiation. The undifferentiated tumor cell differentiates to its original phenotype, melanocyte, osteocyte or chondrocyte. Once it has entered the differentiation pathway it cannot turn back. Importantly, the production of infectious and transforming RSV was not influenced by the cell phenotype.
Basically similar experiments were performed on a chicken erythroleukemia system, driven by the temperature sensitive v-erbB oncogene6-8 and by Jain et al. on tamoxifen regulatable myc.9 The temperature sensitive v-erb B was switched off at the non-permissive temperature. Within half an hour or less the cells entered the irreversible pathway of terminal differentiation, as signaled by the appearance of DNase hypersensitive sites around the globin gene. The consequences of the myc switch-off depended on the cell type. The transformed osteoblasts differentiated terminally into bone, and the thymus went to apoptosis. The tissue context and/or the normal development program of the cell thus plays an important role in determining the fate of the de-blocked cell.
Statement 2. Physiological or artificial inducers of differentiation may revert the malignant phenotype. This field started with the demonstration that mouse leukemia could be induced to differentiate terminally by normal cytokines or by artificial agents like TPA or DMSO.10 The pioneer of the field, Leo Sachs, stated that he could “cure” any mouse leukemia by the appropriate differentiation inducing treatment. It did not work equally well on human leukemia. Nevertheless, proof of principle was provided by some established human leukemia lines. The erythroblastic leukemia line K562 could be induced to differentiate terminally by TPA or DMSO.11 The myeloid leukemia line HL60 could be induced to differentiate into a macrophage or granulocyte, depending on the inducing agent.12,13
Differentiation is the mechanism for the remarkable cure of human PML that carries a translocation between the PML oncoprotein and the RAR-α retinoic acid receptor. Their fusion product inhibits retinoic acid-induced granulocytic differentiation in human myeloid cells.14,15 Induction of differentiation may also play a role in the protection of the host against potentially dangerous viruses. Epstein–Barr virus (EBV) is a case in point. It is the most highly transforming known virus in relation to the most susceptible host cell, the human B-lymphocyte. We have lived with this virus for millions of years, because all Old World (but not New World) primates carry viruses of the EBV family. In genetically (e.g., Wiskott-Aldrich) iatrogenically (transplant recipients) or infection-(HIV) induced immunosuppression, as well as in the immunologically naive New World primates, primary EBV infection can induce fatal lymphoproliferative disease. Our recent work shows that immunocompetent humans are protected not only by the cytotoxicity of CD8+ T-cells, seen only in seropositives, but also by CD4+ T-cells in both seropositives and seronegatives (innate immunity).16
Statement 3: Fully fledged tumor cells with the required genetic and epigenetic changes often fail to grow, due to the inhibition by the microenvironment. The same applies to the vast majority of cancer cells that disseminate throughout the body of cancer patients but never grow into metastases.17 Part of this suppressive effect is mediated by stroma tumor cell interactions.
Statement 4: In or near the successfully proliferating tumors the originally inhibitory stroma becomes “corrupted” so that it loses its inhibitory potential and may even become stimulatory.18 Understanding the nature of this corruption in mechanistic terms may have therapeutic implications toward tightening the surveillance of the patient.
Statement 5: Microenvironmental control is multicomponental and involves different players. This can be exemplified by the study of Partanen et al.19 The forward propelling force of the potential tumor cell—or its quiescence—is primarily determined by compounded changes at the cellular level. Oncogene activation initiates the drive. The induction of basal membrane function by Matrigel in 3D cultures of immortalized but non-neoplastic mammary epithelium leads to acinus formation. This may inhibit the forward propelling effect of the cellular changes. There is no such inhibition in 3D cultures in collagen where acini are not formed. Acinus formation may be blocked and Matrigel action prevented by knocking out the polarity gene LkB1.
No potential conflicts of interest were disclosed.