J. Michael Bishop (2014) - Forging a Genetic Paradigm for Cancer

Thank you very much, it's been 25 years since Harold Varmus and I received the Nobel prize and I'm embarrassed to admit this, this is my first appearance at the Lindau meeting. I certainly confess I should have come sooner but on other hand the lengthy interval allows the opportunity to look back and see not only what the foundation might be for a Nobel discovery but the extraordinary ramifications that it can have over the ensuing quarter of a century. In 1966 at the age of 85 Peyton Rous received the Nobel Prize for his discovery of a virus that causes sarcomas in chickens, work that he had performed 55 years earlier. He opened his Nobel lecture with the following comments: As flesh of his own flesh which is somehow been rendered proliferative, rampant, predatory and ungovernable. Tumours are the most concrete and formidable of human maladies. Yet despite more than 70 years of experimental study they remain the least understood. What can be the why for these happenings?" Well we now know the why for these happenings: cancer arises from the malfunction of genes. The malfunction takes 2 forms. Gain of function, effecting genes known collectively as proto-oncogenes and loss of function affecting genes known as tumour suppressor genes. My purpose today is to sketch the history of how this paradigm took shape and where it has led us. It is a story that's really rich with illustrations of how science proceeds. Now the genetic paradigm for cancer has been more than a century in the making and a number of clues set the stage. First, and the earliest and really most fundamental, was the fact that malignancy is a durable cellular phenotype. Rudolf Virchow is a 19th century pathologist and polymath who took note that the cells of multiple metastases in an individual all resemble one another as well as the cells in the primary tumour. He concluded that all these cells might have had a common origin, which in turn implied a large number of cellular divisions throughout which the malignant phenotype had been preserved. This line of thinking persuaded Virchow that the newly emerging and still very controversial theory of the cell was correct: all cells indeed come from cells. And he became the most ardent and effective advocate for that theory. And now consider HeLa cells. In 1952 Henrietta Lacks was treated for cervical carcinoma at John Hopkins University. And without her knowledge cells from her tumour were placed into tissue culture where they have thrived to this day and are widely used in medical research around the globe. Someone who must have had some time on their hands has computed that the cells have now grown to an accumulate bulk of 20 tons. They remain highly malignant and they seem to be almost everywhere. An astonishing number of allegedly unique cell lines are in reality contaminated with or composed entirely of HeLa cells Henrietta Lacks and her immortal remnants have become a coloss celeb in the United States because of a book about them that has now been in the New York Times best seller list for more than a hundred and sixty weeks. Conclusion, the durability of the malignanat phenotype has to suggest a genetic underpinning. The second clue was the discovery that many causes of cancer are in fact directly mutagenic. The existence of external agents that cause cancer was established gradually over the course of more than two centuries. The culprits included both chemicals and radiation. The initial evidence was limited to guilt by association for example individuals who spent an exceptional amount of time in the sun were unusually prone to skin cancer. In due course however experimentalist wade in. the 4th developed a highly invasive sarcoma at the side of the radiation after a hiatus of more than 12 months. However frail the experiment now seems to us, it is still viewed as the first induction of cancer in an experimental animal. Clunet was inspired to do the experiment of course by the occurrence of malignant skin cancer in the early researchers with X-rays, whom he described as I quote "victims of their own radiological imprudence". And then Ernest Kennaway in Britain refined that discovery by purifying a carcinogenic agent from coal tar, identifying it as benzanthracine and then synthesising benzanthracine from scratch and showing that it was indeed carcinogenic. So the discovery of external carcinogens immediately raised the question of, How do they work? The first insight came from H. J. Muller when he demonstrated that X-rays caused mutations in fruit flies. He concluded his first report of this finding, the very last sentence, with a suggestion that mutation might explain the cancers elicited bioradiation - the thought fell on deaf ears. Within a decade, however, the carcinogen Methylcholanthrene was shown to be mutagenic in mice and then in the early 1970 Bruce Ames and others demonstrated that many chemical carcinogens are mutagenic when applied to bacteria or eukaryotic cells. When all is said and done at that point, however, none of this produced a consensus that mutation of genes might be responsible for cancer. Meanwhile a third clue was making its self known. Cancer cells have both numerical and structural abnormalities of chromosomes. It has become essential to begin this story with the German biologist Theodor Boveri. Inspired by observations that he made with fertilised eggs sea urchins and worms Boveri joined William Sutton in the United States as the first to recognise that chromosomes are the physical carriers of heredity. And then in a stunning feat of imagination, without ever laying hands on a cancer cell, Boveri conceived the idea that gain and loss of chromosomes might be responsible for cancer. A circumstance we now call aneuploidy and we know to be common in cancer cells. He developed this idea in a monograph that was published in 1914. It went unappreciated at the time, it is now considered a classic to the biological literature, and that's my own prized copy up there on the screen. Boveri's theoretical incrimination of chromosomes in cancer lay fallow until 1961. When the pathologist Peter Nowell recruited the graduate student David Hungerford to help him use the nascent technique of kariotyping to explore the chromosomes of leukaemia. They uncovered a miniature chromosome that was consistently present in the cells of chronic myeloid leukaemia: the Philadelphia chromosome, named for the city in which it was discovered. Janet Rowley would eventually show that this miniature chromosome is the product of what we call a reciprocal translocation between chromosomes 9 and 22. The two chromosomes swap pieces of themselves. Hungerford was a heavy smoker who died prematurely of lung cancer. But Nowell and Rowley lived to see their remarkable discovery give rise to a dramatic therapeutic advance: the wonder drug Glivec which has literally transformed the treatment of the disease in which the Philadelphia chromosome was first discovered. Gradually translocations in other forms of leukaemia came into view but for quite a while chromosomal translocations were thought to be exclusive to leukaemias. That proved to be a misapprehension created by the relative difficulty of karyotyping solid tumours. Having mitigated that difficulty we now know that abnormalities of both chromosomal number and structure are prevalent in virtually all malignancies, as exemplified by this kariotype of colon cancer. It is sheer mayhem and it is easy to imagine that this might disturb the behaviour of a cell. The fourth clue strikes at the heart of the matter: Cancer is sometimes heritable. In 1866 the neurosurgeon and anthropologist Paul Broca took time out to sketch the medical pedigree of his wife's family and found himself confronted with an inherited predisposition to breast cancer. Or perhaps alternatively to an environmental effect. Broca went on to expand this observation by identifying multiple human pedigrees in which cancer seemed to be inherited. It was an insight far ahead of its time. We now know that perhaps 10 percent of human cancer is familial. Inherited in a Mendelian dominant manner but paradoxically usually transmitted by a recessant genetic lesion, and we'll talk a little bit about that later. The tumour type in cancer families is determined by the responsible genetic lesion. In this family the inherited disease is inevitably breast cancer. But here in another family pedigree it's riddled with diverse forms of cancer. The explanation for these differences lies mainly in the nature of the responsible gene. The well known BRCA1 gene in the case of the breast cancer family and the gene known as GP53 in the case of this family with multiple kinds of cancer. The truth is we don't fully understand the tissue specificity. Now a fifth clue was provided by the discover that inherited defects in DNA repair created a predisposition to cancer. The first example came from the study of patients with Xeroderma Pigmentosa who inherited a horrendous predisposition to skin cancer elicited by sunlight. Witness this unfortunate Israeli patient. In 1968 at UCSF the year I moved there, James Cleaver reported that individuals with Xeroderma Pigmentosa had inherited a defect in the machinery that normally repairs the DNA damage inflicted by ultraviolet light. It was an easy leap to suspect that unrepaired DNA might be a seed bed for cancer. In the ensuing years other defects in the response to DNA damage had been implicated in both inheritable and what we call sematic non-inherited cancer. These diverse clues were all at hand by the late 1960s. But many still turned a blind eye, Peyton Rous among them. Rous was not enamoured of the genetic explanation for cancer to say the least. He put his distaste for the explanation into his Nobel lecture. Here's another quote - somatic mutations as these are termed. But numerous facts, when taken together, decisively exclude this possibility. This explanation acts as a tranquiliser on those who believe it." Rous died in 1970, just in time to avoid the embarrassment of seeing his dogmatism rudely upended. And ironically the upending began with a virus that he had discovered, which brings us to my part of the story. When I began my research career in 1968 I was working on polio virus. But a new found colleague at UCSF Warren Levinson introduced me to Rous sarcoma virus. I really knew nothing about it before. Here is what I saw. Infection of chicken cells with the virus converts them to a facsimile of cancer cells within 24 hours. Put the infected cells into a chicken and they will quickly grow into a lethal tumour. It seemed to me that the virus of Peyton Rous represented a promising handle on the door to the mysteries of cancer. It was a simple and tractable experimental system, whose biological properties were well described: It was capable of rapidly converting cells from normal to cancerous growth. There was a quantitative assay for the virus in-vitro which made the virus amenable to genetic analysis. The virus could be purified in really large quantities facilitating physical and biochemical analysis and it was safe to use. This virus did not infect human cells. At the time Rous sarcoma virus posed 2 great puzzles: First, it appeared that the viral genome was established as a heritable property of the host cell. How could this happen with a virus whose genome is composed of RNA not DNA? Second the virus is capable of rapid transformation as I illustrated. How is that accomplished? The first puzzle was solved in two steps. The discovery of reverse transcriptase which copies the viral RNA into DNA, that occurred in 1970. And the demonstration that the DNA product of reverse transcription is integrated into the chromosomal DNA of the host cell, where it is perpetuated and expressed as an unwelcome addition to the cellular genome. The demonstration of integration was the first research that Harold Varmus performed after joining me as a postdoctoral fellow in 1970 and we were to work together as co-equals for more than a decade. Meanwhile genetic analysis had revealed that Rous sarcoma virus actually possesses a gene, an oncogene, devoted solely to the malignant transformation of cells. The oncogene was dubbed src for the sarcomas that it elicits in chickens. And remarkably src plays no role in viral replication. That is the job of the other genes in this exquisitely compact and efficient genome. So src raised a evolutionary puzzle. If it is irrelevant to viral replication, why is it there? Given the intimate interaction between viral and cellular genomes that Harold and I had studied it occurred to us that Rous sarcoma virus might have acquired src from the cells in which it replicates. Harold and I set about to test this idea. To do so we devised a DNA probe that could detect the src gene with great sensitivity and specificity and then assayed and utilised the probe - neither were available before. It was a laborious effort mainly because we were doing this before the era of recombinant DNA. The requisite experiments occupied more than 2 years. The fact that they were done at all was a tribute to the valiant efforts of two postdoctoral fellows Dominique Stehelin and Ramareddy Guntaka and today the work could probably be done in a matter of weeks. Am I sorry we didn't wait? Absolutely not! By 1976 we had a persuasive answer to our question and it was a game changer. The viral oncogene had indeed been pirated from a host cell. It is a fully spliced version of its progenitor except at some point the cellular gene sustained a mutation converting it to an oncogene which made the virus suddenly very apparent, very obvious. The first sighting of src in normal DNA was truly a rivetting event. Here was how Dominique Staehlin later described his reaction, normal DNA contains sequences related to the src gene of the transforming virus. I suspect that few have had the privilege of enjoying such a moment when one is intensely and profoundly aware that a major step forward in science has been made and that one has contributed to it." It soon became apparent that src was more than an isolated curiosity. The inventory of retroviral oncogenes mounted steadily and each of these had been derived from the genome of a normal cell. Thus for each viral oncogene there was a cognate cellular proto-oncogene. Accidents of nature had uncovered a whole battery of potential cancer genes in normal cells. The cellular absorbed src proto-oncogene is a well behaved and vital switch in the daily affairs of normal cells. The viral src oncogene is a mutant manufacture whose gene product has been constitutively activated as in fact an enzyme, but I have no time to talk about that, a gain of function that creates a cancer gene. Gaining function by one means or another proved true for all the other retroviral oncogenes acquired from normal cells. So once again it was easy to imagine that this battery of cellular proto-oncogenes might represent a keyboard on which all matter of carcinogens might play. Creating cellular oncogenes without the involvement of a virus. So that raised the question, might proto-oncogenes be malfunctioning in human cancer cells? Within the space of one year, between 1982 and 1983, 3 affirmative observations observed. First a point mutation was found in a previously identified proto-oncogene RAS. And this mutation was in the RAS of human tumour cells. The mutation constitutively activates the gene product just as the case with viral src, a gain of function. Mutation in any of the 3 existing RAS genes proved to be common in a variety of human tumour types. Second, chromosomal translocations associated with the disease Burkitt Lymphoma, relocated previously identified oncogene known as Myc, thereby deregulating and augmenting its transcription, gain of function of a different sort. Over-expression of Myc by one means or another has proven to be one of the most frequent genetic malfunctions in human cancers of various sorts. And third, focal amplification of chromosomal ...(inaudible 19:55) in tumour cells were found to contain proto-oncogenes, creating extravagant expression of the genes. Perhaps the most telling example involved a cousin of Myc, designated MYCN, which our research group discovered because it is amplified in a substantial portion of childhood neuroblastomas. We've soon learnt that amplification of MYCN was associated exclusively with the more aggressive forms of neuroblastoma. And this association between a proto-oncogene and tumour phenotype added a new dimension to the circumstantial evidence incriminating proto-oncogenes in human tumours and it also established a powerful biomarker for the clinical management of neuroblastoma that is widely used to this day. If a child's tumour has amplified NMYC, it is not going to respond to conventional therapy. If it does not, the child is likely to be cured. It is now axiomatic that at least one proto-oncogene has suffered gain of function in virtually all human tumours. With the advent of advanced genomic technologies the inventory of culpable proto-oncogenes has continued to mount far beyond the number uncovered by retroviruses. As the discovery and characterisation of proto-oncogenes proceeded another sort of cancer gene was also coming into view. A hint of something different came from experiments in which cancer cells were fused with diploid normal cells. In some instances the fusions suppress the malignant phenotype. As if the cancer cell might have suffered from a recessive deficiency that can be rectified by the addition of one or more normal chromosomes. Henry Harris was the principal architect of this work and was known to use the term 'tumour suppressor gene' in explaining the results. And that term eventually made its way into the cancer lexicon. However, the results were received with considerable scepticism. The first vindication of Harris, a clear siting of a recessive cancer gene, came from the cytogenetic study of the familial form of childhood retinoblastoma. As illustrated here the disease is inherited in a Mendelian dominant manner but that is a biological illusion. The illusion was unmasked by cytogenetisists who by 1983 had identified a small deletion in chromosome 13 of affected individuals. The deletion segregated with inherited retinoblastoma. And at least some tumours, both alleles of the suspect region, were gone. Thus it appeared that the predisposition of the retinoblastoma was genetically recessive, it might well involved a single gene. The crucial gene removed by the deletion was eventually isolated and christened RB1. In both familial and non-hereditary versions of retinoblastoma inevitably suffer homozygous deficiencies in this gene. Now if the deficiency is recessive, why is the tumour inherited in a dominant manner? Well, children born with a heterozygous deficiency in RB1 will have that deficiency in every retinal cell, a headstart towards tumour genesis. One additional genetic event that incapacitates the remaining normal alleles of the gene in a single cell will launch tumour genesis. The likelihood that this will occur in at least one retinal cell is apparently high, which in turn masks the recessive nature of the genetic lesion. In reality the second event is often sufficiently frequent to engender multiple tumours in both eyes. In the absence of an inherited defective allele both alleles of RB1 must be crippled in the same somatic cell before a tumour can arise. Thus non-hereditary retinoblastoma is exceedingly rare. RB1 became the archetype for what we now call 'tumour suppressor genes'. Homozygous deficiencies in such genes are a ubiquitous feature of human cancer. In fact they are more common, there are more deficiencies of human suppressor genes in the average tumour than there are gains of function for proto-oncogenes. The ability of these genes to deter tumour genesis resides in their normal functions such as control of the cell cycle and maintaining the integrity of the genome. The inventory of these genes, like that of proto-oncogenes, is still growing. Heterozygous deficiencies in one or other tumour suppressor gene are responsible for virtually all inherited cancer. Mutant proto-oncogenes do not figure much in congenital cancer probably because of their genetic dominance. Even a single allele is generally lethal to embryonic development. So we have come to understand tumorigenesis as a very nasty collaboration between 2 sorts of genetic malfunction: gain of function which jams accelerators of the cell proto-oncogenes, and loss of function which cripples brakes of the cell tumour suppresser genes. With these culprits in hand we turn to the question of how they are deployed during tumour genesis. Tumours appear to arise in a stepwise manner orchestrated by natural selection in miniature. Two sorts of observations gave rise to this multi-step view of cancer. First, the pathologists have long known that many tumours display discrete morphological stages in their development, progressing from the benign to the malignant in an incremental manner. As illustrated here with the development of a carcinoma in epithelium. You don't have to be a pathologist to recognise that reading from left to right things get increasingly ugly. Second, the incidents of most cancers accrues over the human life span, as illustrated here for colon cancer. The lengthy time to disease and the exponential nature of the curve are best explained by the accretion of multiple events that sum to the malignant phenotype. The data can be used to estimate the number of these events. The estimates range from 1, in some childhood tumours, to more than a dozen, in prostate cancer. And to a first approximation these numbers derived from epidemiological data are being borne out by genomic analysis. In 1976 building on the supposition that most if not all tumourigenesis originates from 1 or at the most a few cells, Peter Nowell of Philadelphia chromosome fame, proposed a Darwinian scheme in which I quote, allowing sequential selection of more aggressive sublines." Genomic science has now put flesh on the bones of this scheme. First, by identifying genetic and epigenetic events that constitute each step forward in Nowell's scheme for various tumours. In at least some incidents both the nature and the sequence of these events appears to be specific to these tumour types. And second, by showing that some of the purported blind alleys, for example the shaded circles up there, some of these alleged blind alleys in Nowell's evolutionary tree actually lead to independent malignant clones. Minority populations that can lurk undetected unless the dominant clone is either reduced by therapy or the experimentalist meticulously searches them out which is rapidly becoming a routine procedure. The scope of genomic heterogenicity within single tumours can be very large. This is a recent revelation that raises serious challenges to taxonomy and therapeutics. In summary, although there are many causes of cancer they may all act by damaging genes or otherwise disturbing their function. The disturbance takes 2 forms: gain and loss of function that combine to create the malignant phenotype. The cooperation usually involves multiple genes brought into play sequentially by a perverse form of natural selection that does the host organism no favour. To fully understand and exploit the genetic pathogenesis of cancer we need a full inventory of cancer genes for each type of human cancer. This is a vast international project, and fuelled by dramatic advances in technology The results have identified several hundred cancer genes. Defective tumour suppressor genes, as I said before, appear to be more frequent than mutant proto-oncogenes. It appears that every tumour has a distinctive genetic fingerprint. If you look at all the mutations, every tumour is unique genomicly. But each type of tumour is distinguished by a subset of shared cancer genes. Indeed genome sequencing is poised to replace the more traditional means of classifying tumours. Genetic paradigm and the insights that it has produced have strengthened our approach to virtually every aspect of the cancer problem, as I've listed here. From the determination of cause, to the improvement of detection, diagnostics and therapeutics. A mere 20 years ago this list would not have been credible. Now it is a rich reality. In March of 1986 the Nobel laureate Renato Dulbecco published an essay in SCIENCE magazine that represented the first formal call for sequencing cancer genomes. And here in part is what he said - Research in human cancer would receive a major boost from the detailed knowledge of DNA. In one generation we have come a long way in our efforts to understand cancer. The next generation can look forward to exciting new tasks that may lead to a completion of our knowledge about cancer, closing one of the most challenging chapters in biological research." Well we have not yet closed the chapter, we are not even certain how long the chapter may be. But we are turning the pages very rapidly. And perhaps out there among you is a young person who will turn the final page. Thank you for your attention.

J. Michael Bishop (2014)

Forging a Genetic Paradigm for Cancer

J. Michael Bishop (2014)

Forging a Genetic Paradigm for Cancer

Abstract

It is now axiomatic that, no matter what its causes, cancer ultimately arises from the malfunction of genes. A number of clues prefigured this paradigm: the persistence of the malignant phenotype through countless cell divisions; the mutagenicity of various agents that can cause cancer; the presence of chromosomal abnormalities in cancer cells; the occasional instances in which cancer presents as a familial disease; and the predisposition to cancer that accompanies heritable deficiencies in DNA repair. In concert, these findings pointed to an altered genome as the underpinning of cancer. It was the study of RNA tumor viruses, however, that first fingered potentially tumorigenic culprits among the genes of normal cells: “proto-oncogenes” that can be converted to “oncogenes” by genetically dominant gain of function. A variety of genetic anomalies inflict gain of function on proto-oncogenes in human cancer, and there is persuasive evidence that these anomalies contribute to tumorigenesis. Meanwhile, hints emerged that cancer cells might also suffer from recessive deficiencies that contribute to the neoplastic phenotype. The first tangible sighting of such a deficiency came from the discovery of a focal chromosomal deletion that creates a hereditary predisposition to childhood retinoblastoma. Exploration of this deletion uncovered the first of many “tumor suppressor genes,” homozygous deficiencies of which contribute to tumorigenesis. Combinations of wayward proto-oncogenes and defective tumor suppressor genes are present in most, if not all human cancers, having accrued in a stepwise fashion to drive the clonal development and diversification that lead to malignancy – a maladaptive form of Darwinian evolution in miniature. The growing power of genomic science promises to bring us a comprehensive inventory of the genetic maladies in human cancer and the particular tumors in which each of the various maladies may be culpable. The resulting genetic fingerprints should strengthen our approach to virtually every aspect of cancer, including predisposition, cause, pathogenesis, detection, taxonomy, therapy, prognosis and prevention. The forging of the genetic paradigm for cancer provides a powerful example of the unexpected ways in which science can unveil the secrets of nature to the benefit of human health and welfare.

Recommended Readings:
Bishop, J. M. Cancer: The Rise of the Genetic Paradigm. Genes and Development 9:1309-15 (1995)
Strattan, M.R. Exploring the Genomes of Cancer Cells: Progress and Promise. Science 331: 1553-8 (2011)

Content User Level

Beginner  Intermediate  Advanced 

Cite


Specify width: px

Share

COPYRIGHT

Content User Level

Beginner  Intermediate  Advanced 

Cite


Specify width: px

Share

COPYRIGHT


Related Content