Milestones in Cancer NATURE 0

MILESTONES
MILESTONE 1
Observations from a ploughman

“What is it that decides what organs shall suffer be considered to have the same exposure to the in accordance with the Paget hypothesis. Using
a case of disseminated cancer?” This question cancer cells because of similar blood flows. mice, they grafted kidney, ovary and lung tissue
intrigued Stephen Paget, assistant surgeon to This was enough to persuade Paget that under the skin or into the muscle, and showed
the West London hospital and the Metropolitan sites of secondary growths are not a matter that the transplanted tissues established their
hospital, whose self-effacing paper of 1889 of chance, and that some organs provide a own blood supply. They then injected the mice
records his careful analyses of case histories that more fertile environment than others for with melanoma cells. Metastases developed
led to the visionary ‘soil and seed’ hypothesis of the growth of certain metastases. “The best in the grafted lung and ovary tissue but not in
metastasis. work in the pathology of cancer is now done the renal tissue, thereby showing a distinct
“When a plant goes to seed, its seeds are by those who … are studying the nature of preference.
carried in all directions,” he wrote. “But they the seed,” he noted. “They are like scientific Notably, radioactive labelling of the injected
can only live and grow if they fall on con- botanists; and he who turns over the records cells showed that they were equally likely to be
genial soil.” This idea was at odds with one of cases of cancer is only a ploughman, but trapped in the kidney tissue as in either of the
prevalent theory of the time, which stated his observation of the properties of the soil other transplants. So, just landing in a tissue
that cancer cells, having been spread through may also be useful.” is not sufficient for cancer cells to develop a
the body in the blood or lymph, could lodge This proved to be the case and, although secondary tumour; rather, some property of the
in a tissue and persuade the surrounding it languished in the shadows for many years, tissue itself must sustain the new growth. The
cells to grow similarly. However, Paget fol- the seed and soil hypothesis was revived fully idea that cancer cells require some ‘nourish-
lowed the school of thought that all cancer in 1980 by Ian Hart and Isaiah Fidler. By this ment’ from their environment to develop still
cells could continually develop wherever they time, clinical observations had established that motivates research today, with the focus now
settled, but grew only in certain organs that certain organs were, indeed, more susceptible being on unravelling the molecular mechanisms
were somehow predisposed to a secondary to metastasis, even after specific properties of that bring seed and soil together to promote
cancer. the tumour cells and other host factors had been metastases.
Paget reasoned that if the organs where accounted for. Helen Dell, Nature, Locum Associate
secondary tumours arose were ‘passive’ in the So, Hart and Fidler examined whether the News and Views Editor
process, then these cancers would be distributed locations of metastases exist merely because References and links
randomly. By analysing 735 case histories of fatal tumour cells tend to come to rest in particular ORIGINAL RESEARCH PAPERS Paget, S. The distribution of
breast cancer, he found that metastases formed organs — for instance, because the blood capil- secondary growths in cancer of the breast. Lancet 1, 571–573
(1889) | Hart, I. R. & Fidler, I. J. Role of organ selectivity in the
in the liver far more often than in any other laries are more narrow — or because the dis- determination of metastatic patterns of B16 melanoma. Cancer
organ — even those such as the spleen that could tributed cells can only grow at particular sites, Res. 40, 2281–2287 (1980)
NATURE MILESTONES | C ANCER APRIL 2006 | S7

© 2006 Nature Publishing Group
MILESTONES
MILESTONE 2 which he had also discovered and named. By

following the fate of cells with different
chromosomes, he surmised that individual
Lack of principles chromosomes were qualitatively dis-
similar and transmitted different inheritance
The genetic basis of cancer is a cornerstone of factors. He then suggested that aberrant
modern cancer research that began to unravel mitoses led to the unequal distribution
over a century ago. of chromosomes, which, in most cases,
In 1890, David von Hansemann described would be detrimental. Yet, on occasion, a “par- © Neil Smith
in detail the mitotic figures of 13 different ticular, incorrect combination of chromosomes”
carcinoma samples. In every case, he found would generate a malignant cell endowed with With astonishing prescience, Boveri applied
examples of aberrant mitotic figures. These the ability of “schrankenloser Vermehrung” his model further to explain the emergence of
included multipolar mitoses and anaphase (unlimited growth), which would pass the defect different tumour types within one tissue, and
figures that showed asymmetric distribution on to its progeny. The foundations for viewing anticipated the clonal origin of tumours, the
of ‘chromatin loops’ (or chromosomes). He cancer as a genetic disease were laid. allelic loss of recessive chromosome elements,
postulated that these aberrant cell divisions Boveri applied his concept to explain dis- the heritability of cancer susceptibilities, the sim-
were responsible for the decreased or increased parate phenomena linked to cancer, and made ilarity of the steps that initiate tumorigenesis and
chromatin content found in cancer cells. a number of bold and bafflingly accurate pre- those responsible for cancer progression, and the
At the beginning of the twentieth century, the dictions. Today, we can see that he foretold the sensitivity of cancer cells to radiotherapy. All of
zoologist Theodor Boveri pursued this — largely existence of cell-cycle checkpoints (“hemmung- these ideas have since found wide acceptance
ignored — association between aberrant mitoses seinrichtungen”), tumour-suppressor genes and molecular explanations.
and malignant tumours. One of his impor- (“teilungshemmende chromosomen”) and onco- Subsequent work by several investigators
tant innovations was to devise experimental genes (“teilungsfoerdernde chromosomen”). showed that known carcinogens, such as ion-
manipulations of sea urchin eggs that allowed He further envisaged that ‘poisons’ (including izing radiation, acted as mutagens, which fur-
him to induce multipolar mitoses and, therefore, nicotine), radiation, physical insults, pathogens, ther underscored the genetic basis of cancer. A
aberrant chromosome segregation. Boveri, chronic inflammation and tissue repair might consistent chromosomal abnormality that was
for example, found ways to generate cells with all be linked to the development of cancer by found in 1960 in chronic myeloid leukaemia
multiple copies of the centrosome — an organelle indirectly promoting aberrant mitoses or other — the Philadelphia chromosome (see Milestone
that organizes the poles of the mitotic spindle, events that cause chromosome imbalances. 10) — lent further support to this idea.
MILESTONE 3 From these and subsequent studies arose showed that spontaneous tumours were not
the belief, summarized by Harold Hewitt and inherently deficient in tumour antigens, but
Hide and seek colleagues, that naturally arising tumours were

not immunogenic. Moreover, Osias Stutman
instead failed to stimulate an effective immune
response. This failure could be overcome by
had reported in 1974 that athymic mice do vaccination, a strategy that has since been
The immune system has an amazing ability to not have an increased frequency of tumours adopted in numerous clinical trials.
seek out and destroy that which is deemed induced by a chemical carcinogen, implying In a technical feat by Pierre van der
foreign, and generally leaves ‘self ’ alone. Yet, that the concept of immune surveillance Bruggen and colleagues, the Boon group later
tumour cells, thanks to accumulated mutations providing protective immunity was incorrect. reported the first identification of a tumour-
and altered patterns of gene expression, differ Yet, in 1982, enthusiasm for tumour specific antigen recognized by cytolytic T cells
from their normal counterparts. Could the immunology was rekindled by the landmark in humans, reinforcing the idea that tumour
same killing power that eradicates infection be discovery by Aline van Pel and Thierry antigens can elicit a detectable tumour-specific
harnessed to destroy cancer cells — cells that Boon that specific immunity to spontaneous response. Whether that response can induce,
are nevertheless self? tumours could be induced by vaccinating mice or be manipulated to induce, rejection of the
Paul Ehrlich thought so. In 1909, he with mutagenized tumour cells. Their study tumour remains unclear. Yet Robert Schreiber
suggested that, thanks to the immune system, and co-workers, in 2001, prompted renewed
tumour development was usually suppressed. interest in immunosurveillance, showing that
Yet, attempts to target tumours by immunodeficient mice are more susceptible
immunotherapy have been less successful to chemically-induced, as well as spontaneous,
than the Ehrlich hypothesis might predict. tumours. This proves to be a ‘catch 22’,
Richmond Prehn and Joan Main, in 1957, however, for the immunocompetent mouse:
showed that tumours induced by chemical in recognizing cancer, the immune system
carcinogens in mice could stimulate tumour- exerts a selection pressure on a tumour cell
specific responses that were able to reject or immunoediting, resulting in its decreased
those same tumours on challenge. They immunogenicity and eventual escape from
concluded that tumour immunity was induced immune-mediated eradication. More recently,
by antigens unique to the chemically-induced Gerald Willimsky and Thomas Blankenstein
tumour, but found that spontaneously arising suggested that sporadic tumours in mice do
tumours were not rejected when tested in the not lose immunogenicity, but rather induce
same experimental manner. tolerance to evade immune detection. How
S8 | APRIL 2006 www.nature.com/milestones/cancer

Closely linked to the Boveri hypothesis is the
idea of genomic instability driving the accumu- MILESTONE 4
lation of chromosome aberrations and muta-
tions in cancer cells. Work by Robert Schimke
and colleagues showing that cancer cells tend to
amplify genes involved in drug resistance, and
From hens to eternity
that these changes can be unstable, was among
the first evidence of genomic instability in cancer. The hypothesis that viruses can cause They found virus particles in almost all of
Today, the concept has been extended by insights cancer has fallen in and out of fashion the vertebrate species that were examined,
into the mechanisms underlying chromosome since the early 1900s. Former US President and showed that tumour incidence
imbalances, increased mutation rates and other Richard Nixon declared war on cancer in corresponded with viral expression in
forms of genetic instabilities, many of which are 1971, at which time many held the view several murine strains. They consequently
that cancer was caused by infective agents; proposed that normal cells had the capacity
relevant to the development of human cancer
however, it is now known that only a few to activate latent tumour viruses, and that
(see Milestone 22).
cancer types can be directly attributed the spontaneous or induced de-repression
With the advantage of present-day knowl-
to viruses. Despite this, work on RNA of a viral oncogene would lead to cancer.
edge, it is tempting to reinterpret the von tumour viruses (retroviruses) led to many Although their hypothesis that most
Hansemann depiction of the “Prinziplosigkeit important discoveries in cancer research cancers were caused by expression of
als Prinzip der Krebszellen” (lack of principle — not least the discovery of some of the retroviral genes was not strictly correct,
as the principle of cancer cells) as the common first cellular oncogenes. their work did lead to the identification
occurrence of chromosome abnormalities and Peyton Rous is surely the grandfather of of the first retroviral oncogene src, the
genetic instability in cancer. the field. His ground breaking work in this realization that these viral genes were
Barbara Marte, Senior Editor, Nature area began in 1910, when he discovered an derived from functional cellular genes
References and links avian tumour that could be transplanted or proto-oncogenes and, finally, the
ORIGINAL RESEARCH PAPERS von Hansemann, D. Ueber
asymmetrische Zelltheilung in epithel Krebsen und deren to other individuals — the first of its kind. identification of cellular proto-oncogenes
biologische Bedeutung. Virchow’s Arch. Path. Anat. 119, 299 The tumour was a spindle-cell sarcoma as precursors of transforming cancer
(1890) | Boveri, T. Zur Frage der Entstehung Maligner Tumoren that originated in a Plymouth Rock hen.
(Gustav Fisher, Jena, 1914) | Schimke, R. T., Kaufman, R. J.,
genes (see Milestones 15,16 and 17). So,
Alt, F. W. & Kellems, R. F. Gene amplification and drug resistance Rous inoculated bits of this tumour this collection of papers was truly ground
in cultured murine cells. Science 202, 1051–1055 (1978) into the breast and peritoneal cavity of breaking, and paved the way for other
FURTHER READING Balmain, A. Cancer genetics: from Boveri
and Mendel to microarrays. Nature Rev. Cancer 1, 77–82 (2001)
other hens, and found that they could be important discoveries in cancer research.
successfully transferred and propagated Emma Greenwood,
through subsequent transplants. Executive Editor, Oncogene
either model relates to tumour growth in A year later, Rous published another References and links
paper, which took this work a giant step ORIGINAL RESEARCH PAPERS Rous, P. A transmissible
humans remains to be determined. avian neoplasm (sarcoma of the common fowl). J. Exp. Med.
further. He made cell-free filtrates from
While the suggestion by Ehrlich that 12, 696–705 (1910) | Rous, P. Transmission of a malignant
the tumour using various protocols, and new growth by means of a cell-free filtrate. JAMA 56, 198
the immune system restricts the growth of (1911) | Huebner, R. J. & Todaro, G. J. Oncogenes of RNA
found that they were sufficient to induce
most tumours might have been optimistic, tumor viruses as determinants of cancer. Proc. Natl Acad. Sci.
tumour growth. So, a biological agent in USA 64, 1087–1094 (1969)
the findings that immune cells do recognize
the cell-free filtrate could cause tumour
tumours have nonetheless catalysed an development; this agent was subsequently
upswing of enthusiasm in the field of tumour shown to be a virus, and was named after
immunology, and offer encouragement for its discoverer as Rous sarcoma virus (RSV).
immunotherapy approaches as a potential The importance of this finding was not
adjunct to present cancer therapy. fully appreciated for some time, and it was
Alison Farrell, Senior Editor, Nature Medicine only in 1966, at the age of 77, that Rous was
References and links awarded the Nobel Prize for this research.
ORIGINAL RESEARCH PAPERS Ehrlich, P. Über den In 1969, Robert Huebner and George
jetzigen stand der karzinomforschung. Ned. Tijdschr.
Geneeskd. 5, 273–290 (1909) | Prehn, R. T. & Main, J. M.
Todaro reported a series of experiments
Immunity to methylcholanthrene-induced sarcomas. J. Natl. that led to their proposal that “there
Cancer Inst. 18, 769–778 (1957) | Hewitt, H. B., Blake, E. R. exists a unique class of viruses
& Walder, A. S. A critique of the evidence for active host
present in most, and perhaps
defence against cancer, based on personal studies of 27
murine tumours of spontaneous origin. Br. J. Cancer 33, all, vertebrates that plays an
241–259 (1976) | Stutman, O. Tumor development after important etiologic role in
3-methylcholanthrene in immunologically deficient
the development of tumours
athymic-nude mice. Science 183, 534–536 (1974) |
van Pel, A. & Boon T. Protection against a nonimmunogenic in these animals”. Their
mouse leukemia by an immunogenic variant obtained by hypothesis was that C-type retroviruses
mutagenesis. Proc. Natl. Acad. Sci. USA 79, 4718–4722 — of which RSV was the most famous
(1982) | van der Bruggen, P. et al. A gene encoding an
antigen recognized by cytolytic T lymphocytes on a human example — could be vertically transmitted
melanoma. Science 254, 1643–1647 (1991) | Shankaran, V. from animal to progeny animal, and
et al. IFNγ and lymphocytes prevent primary tumour from cell to progeny cell, and that their
development and shape tumour immunogenicity.
Nature 410, 1107–1111 (2001) | Willimsky, G. &
activation by host genetic factors or
Blankenstein, T. Sporadic immunogenic tumours avoid environmental factors results in oncogene
destruction by inducing T-cell tolerance. Nature 437, expression and cell transformation.
141–146 (2005)

MILESTONES
MILESTONE 6
MILESTONE 5
The needle in
The enemy within the haystack
It is now well accepted that hormones influ- Are all cancer cells equal or are some uniquely
ence the initiation and progression of cancer; responsible for initiating and sustaining the
however, it took almost a century of research to growth of a tumour? Although the idea of a can-
move from early observations that hormone- cer stem cell (CSC) or tumour-initiating cell was
ablative surgery benefits some cancer patients well entrenched by the 1960s, it was not until the
to the development of the first drug against an mid-1990s that these cells were identified and
endocrine target. Although hormones are now characterized.
implicated in several cancers, the study of the Most of our understanding of CSCs has
relationship between oestrogen and breast can- They published the first detailed study of the come from studying haematopoietic malignan-
cer has yielded the most important milestones. anti-oestrogenic effects of ICI,46,474 on the cies. Jacob Furth and Morton Kahn were the
Back in 1915, the first suggestion that a hor- reproductive cycle of rats, and found it to be a first to allude to CSC principles in 1937. Using
mone was involved in tumorigenesis came from safer version of known anti-oestrogens. When cell lines, they provided the first quantitative
Abbie Lathrop and Leo Loeb, who reported the the development of ICI,46,474 as a contraceptive assay for the assessment of the frequency of the
influence of internal secretions from the corpus ultimately stalled, Walpole convinced Imperial malignant cell maintaining the haematopoietic
luteum (ovarian follicles) on the development of Chemical Industries to market it for the treat- tumour, at a time when the origin of leukaemia
spontaneous tumours in mice. Their small but ment of breast cancer. Yet clinicians were slow as being viral or cellular was in dispute. They
landmark study showed that tumour incidence to adopt the drug, and it was not until V. Craig showed that a single leukaemic cell was able
was delayed and reduced from 60–90% to 9% in Jordan showed that it could prevent mam- to transmit the systemic disease when trans-
female mice castrated before 6 months of age. mary tumours in mice that they were finally planted into a mouse. This was followed by
As it was already known that the corpus luteum persuaded to undertake the clinical studies that the development of quantitative methods in
secreted an uncharacterized substance that ultimately led to the 1973 approval of ICI,46,474
induced growth of the breast during pregnancy, as tamoxifen.
the authors speculated that this chemical might Still, the value of tamoxifen in preventing MILESTONE 7
be involved in tumour formation. Eight years human breast cancer was not realized until it was
later, Edgar Allen and Edward Doisy identified studied as an adjuvant to breast cancer surgery.
this substance as oestrogen. Following a trial showing that treatment with Bloodlines
Over the next 25 years, the research of tamoxifen after surgery reduced the incidence of
Abraham Lilienfeld, Brian McMahon, Philip contralateral breast cancer, a large-scale Breast In 1939, Gordon Ide and colleagues adapted
Cole and others into the epidemiological rela- Cancer Prevention Trial was started in 1992 a technique to study the growth of blood
tionship between female reproduction and by Bernard Fisher and colleagues to study the vessels around tumour tissue transplanted
into the rabbit ear. Observing robust tumour
breast cancer lent weight to the hypothesis that drug as a chemopreventative agent. The results
growth and induction of a complex vascular
oestrogen was a carcinogen. However, it was were surprisingly positive — tamoxifen caused a
network, they made the seminal suggestion
the discovery of the oestrogen receptor (ER) 50% reduction in the incidence of breast cancer
that tumours might produce a ‘vessel
by Elwood Jensen in 1958, and his pioneering — which supported its use as a prophylactic
growth-stimulating substance’. In 1945,
study in 1971 on the effect of adrenalectomy on drug in high-risk breast cancer patients. Glenn Algire and colleagues furthered these
human breast cancer, that truly revolutionized Since then, tamoxifen has paved the way studies by a detailed kinetic analysis of the
this field. for research into the design of selective hor- vascular response to tumour transplants.
Jensen studied breast cancer patients to mone-modulating drugs for a range of different They postulated that the growth advantage
correlate the level of ER expression on tumours tumour types. The successful development of of a tumour cell over its normal counterpart
with the response to hormone-ablative surgery. these drugs might well be the first of many more might not be owing to “some hypothetical
He found that breast tumours fell into two milestones in this field. capacity for autonomous growth inherent
categories — ER-rich and ER-poor — and Joanna Owens, Senior Editor, within the [tumour] cell,” but rather to its
patients who had tumours with a high level Nature Reviews Drug Discovery ability to continuously induce angiogenesis
of ER expression were more responsive to References and links — that is, the formation of new blood vessels.
hormone-ablative therapy. This led Jensen to ORIGINAL RESEARCH PAPERS Lathrop, A. E. C. & Loeb, L. This insightful conclusion presaged the
Further investigations on the origin of tumors in mice. III. On the
propose that the ER status of a tumour could realization that a tumour would not
part played by internal secretion in the spontaneous
predict the response to therapy. development of tumors. J. Cancer Res. 1, 1–19 (1915) | Jensen, efficaciously grow in the absence of a blood
Although this evidence for the role of the E. V. et al. Estrogen receptors and breast cancer response to supply and, therefore, that inhibiting
adrenalectomy. Natl Cancer Inst. Monogr. 34, 55–70 (1971) | development of the tumour vasculature could
ER in breast tumours raised the possibility of Harper, M. J. K. & Walpole, A. L. A new derivative of
developing anti-oestrogenic cancer drugs, the triphenylethylene: effect on implantation and mode of action in be exploited as a therapeutic strategy.
pharmaceutical industry was instead focusing rats. J. Reprod. Fert. 13, 101–119 (1967) | Fisher, B. et al. In 1968, Melvin Greenblatt and Philippe
Tamoxifen for the prevention of breast cancer: report of the Shubik showed that tumour transplants
on anti-oestrogenic compounds as contracep- national surgical adjuvant breast and bowel project P-1 study.
stimulated the proliferation of blood vessels
tives. One such candidate was ICI,46,474, which J. Natl Cancer Inst. 90, 1371–1388 (1998)
FURTHER READING Jordan, V. C. et al. Tamoxifen: a most even when a physical barrier — a Millipore
was a non-steroidal anti-oestrogen described by unlikely pioneering medicine. Nature Rev. Drug Discov. 2, filter — was placed between the tumour
Michael Harper and Arthur Walpole in 1967. 205–213 (2003)

the 1960s and 1970s to measure the clonogenic provided a reproducible way of enriching cells
potential of the cell type able to sustain tumour with tumour-initiating activity and ruled out
growth in vivo. Robert Bruce and Hugo Van the stochastic model, which predicted that
der Gaag used the spleen colony-forming assay such an activity would be present in every cell
(CFU-S) — a tool first developed by James Till fraction. AML-initiating cells were not only to determine whether such cells arise through
and Ernest McCulloch, and now widely used in able to differentiate and proliferate, but also mutations accrued in normal tissue stem cells
stem-cell biology — to show that only a small had the capacity to self-renew in vivo — a key or whether stem-cell properties are acquired in
subset of primary cancer tissue was able to attribute of stem cells. more differentiated progenitor cells.
proliferate in vivo. Collectively, these studies Recently, studies in solid tumours have Myrto Raftopoulou, Associate Editor,
underscored the functional heterogeneity in revealed that the CSC concept extends beyond Nature Cell Biology
tumours — not every cell is able to proliferate to haematopoietic malignancies. Michael Clarke References and links
ORIGINAL RESEARCH PAPERS Furth, J. & Kahn, M. C. The
form a colony in vitro or to give rise to a tumour and colleagues, and Peter Dirks and co-workers transmission of leukaemia of mice with a single cell. Am J. Cancer
when transplanted in vivo — and introduced the showed that human breast and brain tumours 31, 276–282 (1937) | Bruce, W. R. & Van der Gaag, H.
A quantitative assay for the number of murine lymphoma cells
concept of CSCs. are not homogeneous, but rather contain a small capable of proliferation in vivo. Nature 199, 79–80 (1963) | Becker,
However, it was not until the identification subset of cells that can be prospectively isolated A. J., McCulloch, E. A. & Till, J. E. Cytological demonstration of
the clonal nature of spleen colonies derived from transplanted
and prospective purification of CSCs by John and are able to initiate phenotypically heteroge- mouse marrow cells. Nature 197, 452–454 (1963) | Buick, R. N.,
Dick and colleagues in 1994 that concrete proof neous cancers in vivo. Till, J. E. & McCulloch, E. A. Colony assay for proliferative blast
cells circulating in myeloblastic leukaemia. Lancet 1, 862–863
was provided for a hierarchical (or stem cell) The identification of solid tumour stem (1977) | Lapidot, T. et al. A cell initiating human acute myeloid
model of cancer. Using limiting dilution analy- cells provided researchers with a firm basis on leukaemia after transplantation into SCID mice. Nature 367,
645–648 (1994) | Bonnet, D. & Dick, J. E. Human acute myeloid
sis together with disease-initiation models, which to re-evaluate cancer therapies to target leukaemia is organised as a hierarchy that originates from a
these investigators showed that when isolated and eliminate not only the bulk population of primitive hematopoietic cell. Nature Med. 3, 730–737 (1997) |
Al-Hajj, M. et al. Prospective identification of tumorigenic breast
from acute myeloid leukaemia (AML) patients, tumour cells, but also the rare but potent self- cancer cells. Proc. Natl Acad. Sci. USA 100, 3983–3988 (2003)
only a small fraction of the tumour cells with renewing cells that initiate and sustain can- | Singh, S. K. et al. Identification of human brain tumour initiating
cells. Nature 432, 396–401 (2004)
a characteristic marker signature were able to cers. Efforts are now underway to unravel the FURTHER READING Wang, C. Y. & Dick, J. E. Cancer stem cells:
establish leukaemia in recipient mice. This mechanisms that regulate CSC function, and lessons from leukaemia. Trends Cell Biol. 15, 494–501 (2005)
and the host stroma. They concluded that article rekindled interest in angiogenesis and bevacizumab was approved by the United
the vessel growth-stimulating substance inspired new investigators to join the field. States Food and Drug Administration in 2004
of Ide and co-workers was a true diffusible Nonetheless, it was another 18 years for the treatment of metastatic colorectal
substance that could, in theory, be identified. before Napoleone Ferrara and colleagues cancer, bringing validation to the Folkman
In 1971, Judah Folkman and colleagues purified and subsequently identified the gene hypothesis of more than 30 years earlier that
isolated just such a ‘tumour angiogenic factor encoding vascular endothelial growth factor targeting the tumour vasculature is a viable
(TAF)’ from tumour extracts, and proposed (VEGF), which is a secreted protein that strategy to treat cancer.
that the growth of malignancies might be can stimulate both vascular endothelial cell Alison Farrell, Senior Editor,
prevented if TAF activity were blocked. proliferation in vitro and angiogenesis in vivo. Nature Medicine
Folkman expanded on this concept in his An isoform of VEGF proved to be identical References and links
ORIGINAL RESEARCH PAPERS Ide, A. G. et al.
visionary synthesis of the contemporary to vascular permeability factor, which was Vascularization of the brown Pearce rabbit epithelioma
tumour-angiogenesis literature, and cloned simultaneously by Pamela Keck and transplant as seen in the transparent ear chamber. Am. J.
proposed that tumour cells secrete a soluble co-workers, and was originally identified Roentgenol. 42, 891–899 (1939) | Algire, G. H. et al. Vascular
reactions of normal and malignant tissues in vivo. I. Vascular
factor that stimulates the proliferation of by Harold Dvorak and colleagues in 1983. reactions of mice to wounds and to normal and neoplastic
endothelial cells, that these in turn control Soon thereafter, two groups independently transplants. J. Natl Cancer Inst. 6, 73–85 (1945) | Greenblatt,
tumour expansion and, in the absence of new showed that the cells nearest to areas of low M. & Shubik, P. Tumor angiogenesis: transfilter diffusion studies
in the hamster by the transparent chamber technique. J. Natl
vessel growth, that tumours do not increase oxygen (hypoxia) in a tumour, and therefore Cancer Inst. 41, 111–124 (1968) | Folkman, J. et al. Isolation of
beyond 2–3 mm in size, entering instead a in most need of blood vessels, had the highest a tumor factor responsible for angiogenesis. J. Exp. Med. 133,
state of ‘dormancy’. He further speculated expression of VEGF, and that hypoxia could 275–288 (1971) | Folkman, J. Tumor angiogenesis: therapeutic
implications. N. Engl. J. Med. 285, 1182–1186 (1971) | Leung,
that anti-angiogenesis — that is, inhibiting directly induce expression of VEGF in cells in D. W. et al. Vascular endothelial growth factor is a secreted
the recruitment of new blood vessels into culture. Gregg Semenza and colleagues would angiogenic mitogen. Science 246, 1306–1309 (1989) |
a tumour and thereby inducing dormancy, later identify hypoxia-inducing factor 1 Keck, P. J. et al. Vascular permeability factor, an endothelial cell
mitogen related to PDGF. Science 246, 1309–1312 (1989) |
such as by an antibody directed against TAF (HIF1) as the transcription factor responsible Senger, D. R. et al. Tumor cells secrete a vascular permeability
— might be a powerful approach to tumour for VEGF expression under hypoxia. factor that promotes accumulation of ascites fluid. Science
therapy. In the meantime, Ferrara and colleagues 219, 983–985 (1983) | Shweiki, D. et al. Vascular endothelial
growth factor induced by hypoxia may mediate hypoxia-initiated
The proposal by Folkman of targeting the provided definitive evidence that VEGF
angiogenesis. Nature 359, 843–845 (1992) | Plate, K. H. et al.
vasculature rather than the tumour cell itself stimulates tumour angiogenesis and growth Vascular endothelial growth factor is a potential tumour
was a complete departure from conventional in mice by inhibiting its function using a angiogenesis factor in human gliomas in vivo. Nature 359,
845–848 (1992) | Forsythe, J. A. et al. Activation of vascular
therapeutic strategies, and was not initially blocking antibody. This finding paved the way
endothelial growth factor gene transcription by hypoxia-
well received by the oncology community. for the development and clinical application inducible factor 1. Mol. Cell Biol. 16, 4604–4613 (1996) | Kim,
Moreover, angiogenesis was no hotbed of a humanized version of the antibody, K. J. et al. Inhibition of vascular endothelial growth factor-
induced angiogenesis suppresses tumor growth in vivo. Nature
of research at the time — the number of bevacizumab (Avastin). Based on early and
362, 841–844 (1993) | Hurwitz, H. et al. Bevacizumab plus
angiogenesis papers published in 1971 could encouraging success in cancer patients when irinotecan, fluorouracil, and leucovorin for metastatic colorectal
be counted on just one hand. Yet, the Folkman used in conjunction with chemotherapy, cancer. N. Engl. J. Med. 350, 2335–2342 (2004)

MILESTONE 9
“This was, I believe, the first
unequivocally-established cause
of cancer that affected large
proportions of the general
It takes (at least)
population.”
Laurence N. Kolonel
two to tango
During the first half of the twentieth century, the
view that genetic mutations could cause cancer
MILESTONE 8 the smoking habits of doctors who lived in large was gaining firm ground (see Milestone 2). Yet,
towns and those who lived in other districts, so ideas about viral causes were also widespread,
the authors concluded that lung cancer could and the genetic model had yet to explain the age
Smoking gun not be attributed to a differential exposure to
atmospheric pollution.
distribution of cancer incidence.
A multitude of studies in the 1950s and
These reports, along with other cohort 1960s applied mathematical models to cancer-
During the first half of the twentieth century, studies published in the 1950s, formed the mortality rates as a function of age, to explore
consumption of manufactured cigarettes basis for the 1964 report of the United States whether a series of mutations, accumulating
increased greatly in the Western world. A Surgeon General, which concluded that over time, could explain the epidemiologi-
rapid increase in lung cancer in men was also “Cigarette smoking is causally related to lung cal data. If cancer was caused by successive
evident, and the prevailing view was that this cancer in men; the magnitude of the effect mutations (or hits) its incidence should be
was a result of improved diagnosis, although of cigarette smoking far outweighs all other associated with a certain power of age. This
there was also discussion about the role of factors.” approach led, for instance, Carl Nordling to
increased air pollution or cigarette smok- The Doll and Hill study is unique in its conclude that, indeed, around seven mutations
ing. More than anyone else, the research of regular updating of the smoking habits of the fitted the age distribution for a range of human
the British epidemiologists Richard Doll and participants. The latest (and final) 50-year cancers. Peter Armitage and Richard Doll
Tony Bradford Hill was responsible for the now follow-up report was published in 2004 by
widely accepted view that most lung cancers Doll and colleagues, including Richard Peto, a
are caused by cigarette smoking. 30-year collaborator on the study. The results
MILESTONE 10
In 1939, a German study indicated that showed that among men born between 1910
non-smokers were more common in healthy and 1930, prolonged cigarette smoking caused
populations than among lung cancer patients.
There followed reports of several case–control
death to occur on average 10 years earlier than
that of lifelong non-smokers, but cessation Cutting and pasting
studies associating lung cancer with cigarette at age 50 halved the hazard and cessation at
smoking, including a study in early 1950 from age 30 almost eliminated it. chromosomes
researchers in the USA, Ernst Wynder and Despite these indisputable data, and con-
Evarts Graham. This study involved over 600 sequent findings identifying the carcinogens A small chromosome identified in the cancer
lung cancer cases and 600 controls. Six months in tobacco and establishing the mechanisms cells of patients with chronic myelogenous
later, Doll and Hill reported a larger case– of carcinogenesis, about 1 billion men world- leukaemia (CML) was the first genetic
control study in the British Medical Journal, wide are daily smokers and smoking still defect to be associated with cancer. This
and concluded that smoking was “a cause, and causes about 1.2 million deaths worldwide chromosome was subsequently found
an important cause” of lung cancer. annually. to be the product of a translocation, a
The real milestone came when Doll and Ezzie Hutchinson, Chief Editor, breakthrough that led to the identification of
Nature Reviews Cancer chromosome translocations in other cancers
Hill designed a prospective cohort study to
and the discovery of many oncogenes.
overcome concerns regarding bias. They sent References and Links
ORIGINAL RESEARCH PAPERS Wynder, E. L. & Graham, E. A. In a paper presented at a National Academy
out questionnaires to more than 34,000 male
Tobacco smoking as a possible etiologic factor in bronchiogenic of Sciences meeting in 1960, University
British physicians to collect details of their carcinoma. JAMA 143, 329–336 (1950) | Doll, R. & Hill, A. B. of Pennsylvania researchers Peter Nowell
smoking habits, which were followed up with Smoking and carcinoma of the lung: preliminary report. BMJ
and David Hungerford first reported the
221, 739–748 (1950) | Doll, R. & Hill, A. B. The mortality of
further questionnaires, and recorded the causes doctors in relation to their smoking habits: a preliminary report. presence of a “minute chromosome” that
of death. The first report of this study was pub- BMJ 228, 1451–1455 (1954) | Doll, R. & Hill, A. B. Lung replaced one of the four smallest autosomes
lished in the British Medical Journal in 1954, cancer and other causes of death in relation to smoking: a
in CML cells. They did not observe this
second report on the mortality of British doctors. BMJ 233,
with a follow-up report in 1956. Their earlier 1071–1076 (1956) defect in other types of leukaemia or
case–control findings were confirmed. They FURTHER READING Müller, F. H. Tabakmissbrauch und lymphoma cells, and many other cells from
lungencarcinoma. Z. Krebsforsch. 49, 57–85 (1939) | United
showed a higher mortality in smokers than these patients contained a normal karyotype.
States Public Health Service. Smoking and Health. Report of the
in non-smokers, and a clear dose–response Advisory Committee to the Surgeon General of the Public Health This unique chromosome was designated
relationship between the amount smoked and Service (United States Department of Health, Education, and ‘The Philadelphia chromosome’ after the city
Welfare, Public Health Service, Center for Disease Control,
the death rate from lung cancer. The data also Washington DC, 1964) <http://www.cdc.gov/tobacco/sgr/sgr_
in which the discovery was made.
indicated a progressive and significant reduc- 1964/sgr64.htm> | Doll, R., Peto, R., Boreham, J. & Sutherland, I. For many years, scientists thought that
tion in mortality with the increase in the length Mortality in relation to smoking: 50 years’ observations on male the Philadelphia chromosome resulted
British doctors. BMJ 328, 1519–1528 (2004) | Hecht, S.
of time over which smoking had been given up. simply from the loss of genetic material.
Tobacco carcinogens, their biomarkers and tobacco-induced
There was remarkably little difference between cancer. Nature Rev. Cancer 3, 733–744 (2003)

reached similar conclusions in a paper from showed that the distribution observed was con-
1954, but, in a further study in 1957, revised sistent with retinoblastoma being caused by two
their model to conclude that the epidemiological mutations. In familial cases, one hit was inher-
data were consistent with many common forms ited whereas the other one was acquired later; in
of cancer developing in two steps, of which one sporadic tumours, both changes were somatic,
or both could be somatic mutations. with a similar mutation rate for both hits. The
Other work in the 1950s and 1960s, includ- Knudson model explained why multiple tumours
ing that of James Neel and Philip Burch, concen- occurred in both eyes in inherited cases, but only
trated on childhood cancers whose development unilaterally in sporadic occurrences. are required for the step-wise progression of
in early life reduced some of the complexity of Knudson and colleagues subsequently tumour development (see Milestone 14).
other forms of cancer; this led researchers to extended the two-hit model to secondary Barbara Marte, Senior Editor, Nature
deduce that multiple, perhaps as few as two, tumours in retinoblastoma patients and to References and links
inherited and/or somatic mutations had a role other childhood cancers. The now famous two- ORIGINAL RESEARCH PAPERS Nordling, C. O. A new
theory on cancer-inducing mechanism. Br. J. Cancer 7, 68–72
in retinoblastoma, neurofibromatosis and child- hit hypothesis was, in later years, to merge with (1953) | Armitage, P. & Doll, R. The age distribution of cancer
hood leukaemias (see also Further Reading). the concept of allelic loss of tumour-suppressor and a multi-stage theory of carcinogenesis. Br. J. Cancer 8,
1–12 (1954) | Armitage, P. & Doll, R. A two-stage theory of
In a seminal paper in 1971, Alfred Knudson genes when it became clear that the develop- carcinogenesis in relation to the age distribution of human
took the idea of multiple hits an important step ment of retinoblastoma was associated with cancer. Br. J. Cancer 11, 161–169 (1957) | Burch, P. R. J.
further. He noted that “what is lacking is direct mutations in both alleles of the retinoblastoma A biological principle and its converse: some implications for
carcinogenesis. Nature 195, 241–243 (1962) | Falls, H. F. & Neel,
evidence that cancer can ever arise in as few gene RB (also known as RB1), and that one RB J. V. Genetics of retinoblastoma. AMA Arch. Ophthalmol. 46,
as two steps and that each step can occur at a mutation was inherited in familial cases of the 367–389 (1951) | Crowe, F. W., Schull, W. J. & Neel, J. V.
A Clinical, Pathological and Genetic Study of Multiple
rate that is compatible with accepted values for disease (see Milestone 11). Neurofibromatosis (Charles C. Thomas, Springfield, IL, 1956) |
mutation rates”. Knudson analysed 48 cases of The current view of cancer has built on these Knudson, A. G. Jr. Mutation and cancer: statistical study of
retinoblastoma for the occurrence of bilateral or findings: we now know that all human cancers retinoblastoma. Proc. Natl Acad. Sci. USA 68, 820–823 (1971)
FURTHER READING Kern, S. E. Whose hypothesis?
unilateral tumours, and the presence of a family display a multitude of genetic and epigenetic Ciphering, sectorials, D lesions, freckles and the operation of
history of the disease. Using Poisson statistics, he changes, and that a number of such alterations Stigler’s Law. Cancer Biol. Ther. 1, 571–581 (2002)
leukaemia cells. She went on to discover Cloning of the breakpoints of other

more than a dozen translocations that were cancer-associated translocations would
specific to other types of cancer cells, notably subsequently lead to the discovery of many
t(15;17) in acute promyelocytic leukaemia other oncogenes, such as B-cell lymphoma
and t(14;18) in lymphoma, which was also 2 (BCL2) and tumour suppressor genes.
described by Carlo Croce and colleagues. Kristine Novak
Cytogenetic analysis is still one of the References and links
ORIGINAL RESEARCH PAPERS Nowell, P. C. & Hungerford,
most reliable methods of diagnosis and
D. A. A minute chromosome in human chronic granulocytic
of determining prognosis in patients with leukemia. Science 132, 1488–1501 (1960) | Rowley, J. D. A new
leukaemia or lymphoma. consistent chromosomal abnormality in chronic myelogenous
leukaemia identified by quinacrine fluorescence and Giemsa
So, how do these chromosome rearrange- staining. Nature 243, 290–293 (1973) | Rowley, J. D.
ments cause cancer? In 1982, Annelies de Identification of a translocation with quinacrine fluorescence in a
Klein and colleagues reported that the human patient with acute leukemia. Ann. Genet. 16, 109–112 (1973) |
de Klein, A. et al. A cellular oncogene is translocated to the
cell homologue (c-ABL; also known as ABL1) Philadelphia chromosome in chronic myelocytic leukaemia.
In the early 1970s, the development of of the transforming sequence of Abelson Nature 300, 765–767 (1982) | Taub, R. et al. Translocation of the
c-myc gene into the immunoglobin heavy chain locus in human
techniques such as quinacrine fluorescence murine leukaemia virus was translocated from Burkitt lymphoma and murine plasmacytoma cells. Proc. Natl
and Giemsa banding allowed researchers chromosome 9 to chromosome 22q in CML Acad. Sci. USA 79, 7837–7841 (1982) | Dalla-Favera, R. et al.
to identify and track segments of cells. This finding indicated a role for c-ABL Human c-myc onc gene is located on the region of
chromosome 8 that is translocated in Burkitt lymphoma cells.
chromosomes. Accelerating the burgeoning in the generation of human leukaemia. At the Proc. Natl Acad. Sci. USA 79, 7824–7827 (1982) | Zech, L.,
field of cytogenetics, Janet Rowley used same time, Rebecca Taub and co-workers, Haglund, U., Nilsson, K. & Klein, G. Characteristic chromosomal
abnormalities in biopsies and lymphoid-cell lines from patients
these technologies to identify a genetic and Riccardo Dalla-Favera and colleagues
with Burkitt and non-Burkitt lymphomas. Int. J. Cancer 17,
abnormality in CML cells — the addition reported the translocation of c-MYC into the 47–56 (1976) | Adams, J. M., Gerondakis, S., Webb, E.,
of extra material to chromosome 9. She immunoglobulin heavy chain locus, through a Corcoran, L. M. & Cory, S. Cellular myc oncogene is altered by
chromosome translocation to an immunoglobulin locus in
then noticed that the amount of additional translocation between chromosomes 8 murine plasmacytomas and is rearranged similarly in human
material was approximately equal to the and 14. This translocation had been identified Burkitt lymphomas. Proc. Natl Acad. Sci. USA 80, 1982–1986
amount missing from chromosome 22, by Laura Zech and co-workers, and is (1983) | Tsujimoto, Y., Finger, L. R., Yunis, J., Nowell, P. C. &
Croce, C. M. Cloning of the chromosome breakpoint of
and proposed that there was a “hitherto frequently observed in Burkitt lymphoma neoplastic B cells with the t(14;18) chromosome translocation.
undetected translocation between the cells. Evidence that the translocation Science 226, 1097–1099 (1984)
long arm of 22 and the long arm of 9” that and subsequent deregulation of MYC FURTHER READING Rowley, J. D., Golomb, H. M. &
Dougherty, C. 15/17 translocations, a consistent
resulted in formation of the Philadelphia expression is an oncogenic event was later chromosomal change in acute promyelocytic leukemia.
chromosome. In the same year, Rowley provided by the Eµ-Myc mouse model Lancet 1, 549–550 (1977) | Adams, J. M. et al. The c-myc
oncogene driven by immunoglobulin enhancers induces
reported another translocation between generated in 1985 by Jerry Adams and lymphoid malignancy in transgenic mice. Nature 318,
chromosomes 8 and 21 in acute myeloblastic colleagues. 533–538 (1985)

MILESTONE 11
The road less

travelled
Few would argue that the path to scientific for a role of tumour suppressor genes in by chromosome walking their way to a cDNA
discovery is short and simple. The all types of cancer: inherited tumours, he fragment that hybridized to transcripts in
realization that cancer could arise through argued in a theoretical paper, were the result normal tissue, but was aberrantly expressed or
the inactivation of recessive genes — tumour of a germline mutation in regulatory genes deleted in retinoblastomas. This pointed to the
suppressors — is a case in point. that suppressed tumorigenesis, followed by inactivation of RB as being causative for cancer,
Throughout the 1970s and 1980s, the somatic loss of the homologous allele. In a conclusion that was confirmed by Huei-
oncogenes dominated the field of cancer non-heritable cancers, both alleles would be Jen Su Huang and colleagues, who rescued
research, and so the prevailing thought was affected in somatic cells. However, the field the neoplastic phenotype of RB-mutant
that tumours were caused by activating had to wait 10 years to pin this hypothesis to a retinoblastoma cells with wild-type RB.
mutations. The famous two-hit hypothesis molecular locus. The involvement of p53 in cancer was
was also finding increasing support (see Then, Webster Cavanee and colleagues known for 10 years before its true role was
Milestone 9), but lacked insights into the localized the retinoblastoma gene (RB; identified. In 1989, Bert Vogelstein’s group
nature of the hits. also known as RB1) to a small region on identified p53 as the gene uncovered by the
Perhaps the strongest impetus to pursue chromosome 13; they showed that inherited cancer-associated deletions on chromosome
the unorthodox idea of tumour suppressor and sporadic cancers had the same second 17p, and showed that one copy was mutated
genes was provided by Henry Harris and hits, and that these cause homozygosity and the other deleted in colorectal cancers.
colleagues, who observed that normal mouse for mutations at the RB region, thereby Similar to RB, the tumour suppressor
cells were dominant to malignant cells when confirming the allelic-hit hypothesis. By function of p53 was confirmed by showing
the two types were fused in the laboratory. the end of the decade, the first two tumour that it rescued the growth phenotype of
This conceptually simple yet technically suppressor genes — RB and p53 (also known p53-mutant carcinoma cells. If p53 caused
demanding work pierced the first hole in the as TP53) — would be identified. tumours only when both alleles were mutant,
theory that (dominant) oncogenes were the In 1986, Stephen Friend and colleagues then it could not be the proto-oncogene it
general rule. isolated a human cDNA that mapped to the RB was widely regarded to be. Arnold Levine’s
While many scientists had previously region and, importantly, was deleted at least group helped to dispel this misconception
presented support for a model of allelic loss partly in tumours. The next year, two groups further, by showing that the p53 mutations
(see Further Reading), it was David Comings — Wen-Hwa Lee and co-workers, followed by that arose in transformed cells in vitro
who, in 1973, articulated a general framework Yuen-Kai Fung and colleagues — cloned RB were of the same type as that which occurs
Kerr, Wyllie and Currie suggested that, and that impaired cell death, similar to enhanced
MILESTONE 12
unlike necrosis, apoptosis might represent a proliferation, was indeed a key step in tumour
genetically regulated cell-suicide programme, development. In the same year, John Reed and
Death defying and, importantly, they stated: “We should now
like to speculate that hyperplasia might some-
colleagues found that overexpression of BCL2
in an immortalized mouse cell line did not
By the late 1960s, it was recognized that the times result from decreased apoptosis rather induce proliferation or transformation in vitro.
spontaneous loss of tumour cells was an impor- than increased mitosis, although we emphasize Although these cells did produce tumours in
tant component in the growth of tumours that we know of no definitive studies to sup- mice, further mutational events were required.
and, although cell death was the most likely port such a hypothesis.” In 1989, Tim McDonnell, Stanley Korsmeyer
cause, little was known about the mechanisms Importantly, in 1988, David Vaux, Suzanne and colleagues reported that the expression of
involved. John Kerr had already shown that Cory and Jerry Adams showed that expression a BCL2–immunoglobulin fusion protein in B
cells died with a morphology that was distinct of the B-cell lymphoma 2 (BCL2) gene, which cells prolonged their survival — an event that
from necrotic cells, but it was not until the had been identified by others as being translo- this group also showed was tumorigenic.
description of apoptosis in a 1972 review by cated in follicular lymphoma (see Milestone 10), Soon after, other oncogenes, such as the
John Kerr, Andrew Wyllie and Alastair Currie could promote the survival of haematopoietic breakpoint cluster region (BCR)–Abelson leukae-
that specific roles for cell death in cancer devel- cells after the removal of growth factors (Gwyn mia viral oncogene (ABL; also known as ABL1),
opment were proposed. Williams and co-workers were later to show that were shown to suppress apoptosis. Conversely,
these growth factors suppressed apoptosis). Vaux several groups, including those of John
and colleagues also showed that the oncogene Cleveland and Gerard Evan, reported that
Myc cooperated with Bcl2 to produce tumours overexpression of MYC induced apoptosis.
in immunocompromised mice. They suggested Initially, this seemed counterintuitive — why
that BCL2 provided a distinct survival signal would the upregulation of an oncogene associ-
that might contribute to neoplasia by allowing ated with increased proliferation induce cell
a clone to persist until other oncogenes, such death? It was proposed that MYC-induced
as Myc, became activated. This and subsequent apoptosis was part of a tumour suppression
An electron micrograph showing an apoptotic
work provided evidence that cell survival was mechanism. Apoptosis as a mechanism to
haematopoietic cell (Bar = 2 µm) © Nicola McCarthy. regulated independently of cell proliferation, limit tumorigenesis was further supported by

in human cancers — that is, they were MILESTONE 13
inactivating mutations that probably acted
in a dominant–negative manner.
Tumour suppressors and oncogenes Environmental
started out at opposite poles; yet, in just
15 years, the field came full circle with the
realization, as Comings had predicted years
awareness
earlier, that tumour suppressors oppose
the action of transforming genes — a Context is everything for cells and, in addition
mechanistic link that has provided the basis to the importance of finding an appropriate
for all subsequent models of malignancy. blood supply (see Milestone 7), tumorigenic
Tanita Casci, Senior Editor, potential will often only be realized when
Nature Reviews Genetics cells find themselves in a tissue environment
References and links that they can subvert to their advantage. This
ORIGINAL RESEARCH PAPERS Harris, H. Cell fusion and
the analysis of malignancy. Proc. R. Soc. Lond. B Biol. Sci. phenomenon was first clearly noted in the 1970s,
179, 1–20 (1971) | Comings, D. E. A general theory of although it was decades before the cellular basis
carcinogenesis. Proc. Natl Acad. Sci. USA 70, 3324–3328
(1973) | Cavenee, W. K. et al. Expression of recessive alleles
was uncovered. grown in culture, they now became trans-
by chromosomal mechanisms in retinoblastoma. Nature In 1975, Beatrice Mintz and Karl Illmensee formed. So, something about the environment
305, 779–784 (1983) | Friend, S. H. et al. A human DNA asked what would happen if mouse tetracarci- of the embryos was able to block tumorigenesis,
segment with properties of the gene that predisposes to
retinoblastoma and osteosarcoma. Nature 323, 643–646
noma cells were placed in a ‘normal’ environ- despite the presence of v-Src. The following year,
(1986) | Lee, W. H. et al. Human retinoblastoma susceptibility ment. They took the tetracarcinoma cells from her group went on to show that wounding was
gene: cloning, identification, and sequence. Science 235, embryoid bodies in vivo, and injected them into one important influence on the ability of a cell
1394–1399 (1987) | Fung, Y. K. et al. Structural evidence for
the authenticity of the human retinoblastoma gene. Science developing mouse blastocysts. Surprisingly, to succumb to tumorigenesis. When RSV was
236, 1657–1661 (1987) | Huang, H-J. S. et al. Suppression normal mice were born with no evidence injected into a chick wing to produce a local
of the neoplastic phenotype by replacement of the RB gene
in human cancer cells. Science 242, 1563–1566 (1988) |
of tumours. When the authors looked more tumour, a second tumour would only be seen if a
Baker, S. J. et al. Chromosome 17 deletions and p53 gene closely, they found that tumour-derived cells wound was simultaneously induced at a remote
mutations in colorectal carcinomas. Science 244, 217–221 were present in large numbers and contributed site. The Bissell group later found that the factor
(1989) | Finlay, C. A., Hinds, P. W. & Levine A. J. The p53
proto-oncogene can act as a suppressor of transformation.
to several unrelated tissues, the most notable responsible was transforming growth factor-β
Cell 57, 1083–1093 (1989) | Baker, S. J. et al. Suppression being functional spermatozoa. From this, Mintz (TGF-β) — an early and surprising demonstra-
of human colorectal carcinoma cell growth by wild type p53. and Illmensee concluded that the tumour cells tion of the dual action of this cytokine.
Science 249, 912–915 (1990).
FURTHER READING Kern, S. E. Whose hypothesis? were developmentally totipotent and could It is only during the past 10 years that we
Ciphering, sectorials, D lesions, freckles and the operation of revert to normal behaviour in the appropriate have begun to understand the molecular basis
Stigler’s Law. Cancer Biol. Ther. 1, 571–581 (2002)
environment. At the time, they also speculated for how the local tissue environment, and proc-
that the original tumorigenic state might not esses such as inflammation and infection, can
the finding that the tumour suppressor p53 involve a mutation. influence tumorigenic cells. For example, in
induced apoptosis (see Milestone 20). This study was to have a strong influence 1997, Bissell and colleagues showed that block-
These discoveries and many others have on Mina Bissell. Inspired by this mysterious ing integrin function was sufficient to revert the
shown that failure to induce apoptosis produces behaviour of tumour cells, Bissell began to malignant phenotype of human breast cancer
hyperplasia, whereas further mutations are focus her own research on the influence of the cells both in culture and in vivo. Others, includ-
required to produce overt neoplasia. Overall, microenvironment. In 1984, she published a ing the groups of Luis Parada and Harold Moses,
the concept that the inability of a cell to die was study, together with David Dolberg, showing were able to show in mouse models that genetic
potentially tumorigenic revolutionized the way that the ability of Rous sarcoma virus (RSV) to alterations in cells of the tumour microenviron-
in which tumorigenesis was viewed and greatly induce tumours was also context dependent. ment contribute to, and can even be sufficient
influenced treatment strategies. The tumour-inducing behaviour of RSV when for initiating, the development of cancer.
Nicola McCarthy, Senior Editor, injected into the wings of newly hatched chicks Alison Schuldt, Senior Editor, Nature Cell Biology
Nature Reviews Cancer was already known, and the viral gene v-src had References and links
References and links been identified as the sole culprit. ORIGINAL RESEARCH PAPERS Mintz, B. & Illmensee, K.
ORIGINAL RESEARCH PAPERS Kerr, J. F., Wyllie, A. H. & Currie, Normal genetically mosaic mice produced from malignant
A. R. Apoptosis: a basic biological phenomenon with wide- What Bissell found, however, was that if teratocarcinoma cells. Proc. Natl Acad. Sci. 72, 3585–3589
ranging implications in tissue kinetics. Br. J. Cancer 26, 239–257 RSV was injected into 4-day-old embryos, no (1975) | Dolberg, D. S. & Bissell, M. J. Inability of Rous sarcoma
(1972) | Vaux, D. L., Cory, S. & Adams, J. M. Bcl-2 gene promotes virus to cause sarcomas in the avian embryo. Nature 309,
haemopoietic cell survival and cooperates with c-myc to
tumours were produced, despite the spread 552–556 (1984) | Dolberg, D. S., Hollingsworth, R., Hertle, M. &
immortalize pre-B cells. Nature 335, 440–442 (1988) of RSV infection throughout the embryo and Bissell, M. J. Wounding and its role in RSV-mediated tumor
FURTHER READING Reed, J. C., Cuddy, M., Slabiak, T., Croce, formation. Science 230, 676–678 (1985) | Weaver, V. M. et al.
C. M. & Nowell, P. C. Oncogenic potential of bcl-2 demonstrated
active v-Src expression. Furthermore, if the Reversion of the malignant phenotype of human breast cells in
by gene transfer. Nature 336, 259–261 (1988) | McDonnell, T. M. infected embryonic cells were isolated and three-dimensional culture and in vivo by integrin blocking
et al. bcl-2–immunoglobulin transgenic mice demonstrate antibodies. J. Cell Biol. 137, 231–245 (1997) | Sieweke, M. H.,
extended B cell survival and follicular lymphoproliferation. Cell 57, Thompson, N. L., Sporn, M. B. & Bissell, M. J. Mediation of
79–88 (1989) | Williams, G. T., Smith, C. A., Spooncer, E., Dexter, wound-related Rous sarcoma virus tumorigenesis by TGF-β.
T. M. & Taylor, D. R. Haemopoietic colony stimulating factors
“These experiments form the Science 248, 1656–1660 (1990) | Zhu, Y. et al. Neurofibromas in
promote cell survival by suppressing apoptosis. Nature 343, groundwork for our current NF1: Schwann cell origin and role of tumor environment. Science
76–79 (1990) | Askew, D. S., Ashmun, R. A., Simmons, B. C. & 296, 920–922 (2002) | Bhowmick, N. A. et al. TGF-β signaling in
Cleveland, J. L. Constitutive c-myc expression in an IL-3-
understanding that the environment in fibroblasts modulates the oncogenic potential of adjacent
dependent myeloid cell line suppresses cell cycle arrest and which cells are growing can influence epithelia. Science 303, 848–851 (2004)
accelerates apoptosis. Oncogene 6, 1915–1922 (1991) | Evan, FURTHER READING Dvorak, H. F. Tumors: wounds that do not
G. I. et al. Induction of apoptosis in fibroblasts by c-myc protein. the expression of the transformed heal. Similarities between tumor stroma generation and wound
Cell 69, 119–128 (1992) phenotype” Sara Courtneidge healing. N. Engl. J. Med. 315, 1650–1659 (1986)

MILESTONES
MILESTONE 14 Milestone 19), he suggested that, as the result of

acquired genomic instability in the expanding
cell population, rare subvariants endowed with
Step by step an extra selective advantage could emerge.
Sequential rounds of clonal selection would
The possibility that cancer could have a genetic produce tumour-cell populations with more
basis was recognized in the early twentieth aggressive phenotypes. Support for this concept © Vicky Askew
century (see Milestone 2). In addition, evidence came from the observation that advanced recognized that these would be difficult to
for a clonal origin of tumours had emerged, as solid tumours often had a greater degree of identify amongst the multitude of evolutionary
had views of cancer as a multistep process. aneuploidy than early stage lesions, and from steps, and that variations due to different
Leslie Foulds, for example, early on described the discovery of specific chromosomal changes selection pressures were likely.
cancer as a “dynamic process advancing that developed during the clinical progression Subsequent years saw the identification of
through stages that are qualitatively different”, of leukaemias. a number oncogenes and tumour-suppressor
progressing from precancerous stages to Nowell discussed the mechanisms genes that were altered in human cancer. In an
increasingly invasive and metastatic stages. underlying genomic instability, such as DNA- influential paper in 1990, Eric Fearon and Bert
Yet, the now prevailing concept of Darwinian repair defects or mitotic errors (see Milestones Vogelstein amalgamated these findings together
evolution and the stepwise progression of 2 and 22), and noted that diverse agents that with the idea of clonal evolution into a coherent
tumours was perhaps most convincingly cause cancer, such as ionizing radiation and molecular model of multistep tumorigenesis.
articulated by Peter Nowell in 1976. His article viruses, can induce genetic changes and might Focusing on colorectal cancer, the authors
incorporated the idea of cancer being caused by contribute to the initial changes as well as the noted the clonal nature of the disease, and
multiple mutations or ‘hits’ (see Milestone 9) into subsequent alterations. the consistent occurrence of mutations in
a general framework of tumour development Nowell wrote “it would be helpful if we could the KRAS oncogene and the allelic loss of
and progression, through the accumulation and associate specific chromosomal alterations known or candidate tumour-suppressor genes,
selection of genetic changes. with particular aspects of tumour suppression”. including p53 (also known as TP53). Although
Nowell concluded that the first step results However, at the time, few consistent changes certain changes were preferentially associated
in cell proliferation that is “unrestrained to had been noted, with the exception of the with specific stages of disease progression,
some degree”, allowing for a selective growth famous Philadelphia chromosome (see the authors documented a multitude of
advantage. While also acknowledging the Milestone 10). Although Nowell anticipated chromosomal and other changes, such as
potential role of epigenetic alterations (see similarities between different tumours, he also frequent DNA hypomethylation of specific
MILESTONE 15 chicken cells, and to less-conserved sequences

in the genomes of other avian species. So,
RSV had acquired its transforming activity
Bad seeds by recombining with, and transducing, the
chicken cellular ‘c-src’ oncogene.
An analysis of the thermostability of the
The identification by G. S. (Steve) Martin DNA duplexes formed between the RSV v-src
of the Rous strain of avian sarcoma virus probe and avian cellular DNAs provided an indi-
(RSV) that was temperature sensitive for cation of the extent of relatedness, and showed
transformation implied that RSV contained that cellular c-src DNA sequences diverged in
an ‘oncogene’ that conferred its tumori- accordance with the phylogenetic distances
genic properties in chickens. Work by Peter between the different species. Using less-
Duesberg and Peter Vogt soon showed that stringent hybridization conditions, Deborah
© Vicky Askew
the genome of RSV contained RNA sequences Spector, Varmus and Bishop detected more
that were missing from replication-competent distantly related c-src sequences in human The 60-kDa proteins shared several antigenic
but transformation-defective viral variants. and mouse DNA, but not in sea urchin, fruit determinants with the viral protein, and were
Using elegant techniques developed in the fly or Escherichia coli. chemically and structurally similar, although
laboratories of Michael Bishop and Harold Bishop, Varmus and colleagues went on not identical. In addition, the cellular homo-
Varmus, Dominique Stehelin used subtrac- to study the gene product of v-src, which logues seemed to function as protein kinases
tion hybridization procedures to generate a had first been identified by Joan Brugge and in a similar way to that previously shown for
cDNA probe that specifically hybridized to Raymond Erikson. Using the same experi- the viral protein (see Milestone 16).
the putative oncogene (or src sequences) of mental approach, the Oppermann, Bishop and Together, these papers provided the first
RSV. In 1976, Stehelin and colleagues found Varmus team (followed a few months later by evidence of the presence of genes related to
that radiolabelled ‘cDNAsrc’ hybridized to the Erikson laboratory) identified the cellular viral oncogenes in the genomes of healthy
homologous sequences in the DNA of normal form by precipitating proteins from extracts vertebrates — this proved, as Bishop famously
of uninfected vertebrate cells with antisera said, that ‘the seeds of cancer are within us’.
derived from rabbits with RSV-induced Their function in the host organism remained
“These were very unexpected findings tumours. This led to the isolation of a 60-kDa unclear, together with the question of whether
that launched the whole field of phosphoprotein in chicken cells, and subse- their function had been altered by the virus.
oncogenes.” Julian Downward quently in quail, rat and human fibroblasts. These discoveries led to an explosion of

MILESTONE 16
regions. They therefore considered the total
accumulation of changes, rather than their
sequence, as most important for tumour
progression. They also concluded that five or
Phosphorylate and conquer
more genetic alterations were probably required
for the development of carcinomas, with fewer
changes needed for benign tumorigenesis. The mystery of the cell-transformation process The association of both virus-induced and
This model of cancer evolution through and the seemingly simple principles that are EGF-induced cell growth with the activation
the accumulation of mutations in both involved in cancerous growth began to be of protein kinases specific for tyrosine was
oncogenes and tumour-suppressor genes, unravelled with the study of retroviruses. certainly intriguing, and indicated a functional
and the stepwise selection of more malignant Research into the function of the protein relationship between oncoprotein activity and
tumour-cell populations, has since been widely products that are encoded by oncogenes receptor signalling.
adopted and generalized to all common forms followed closely on the heels of the discovery of Importantly, shortly thereafter in 1983,
of cancer. At a time when we are beginning to the oncogenes themselves (see Milestones 15 Mike Waterfield and colleagues, and Harry
see the basic understanding of the molecular and 17). Antoniades and colleagues identified the
changes underlying tumorigenesis translated In 1978, in an effort to understand the putative transforming protein of simian
into the development of targeted therapies function of the src oncogene from the avian sarcoma virus as platelet-derived growth factor
(see Milestone 24), it is well worth noting Rous sarcoma virus (RSV), Michael Bishop (PDGF). In addition, in 1984, Mike Waterfield,
the foresight of Nowell in suggesting that and colleagues prepared anti-sera from Yossi Schlessinger and colleagues reported that
individual differences in the genetic and rabbits that were infected with RSV-induced v-erbB from the avian erythroblastosis virus
biological changes in each tumour might tumours, to isolate and further characterize encodes a similar protein to EGFR. These three
warrant personalized therapies. the putative product of src. They found papers, along with those identifying the cellular
Barbara Marte, Senior Editor, Nature that the anti-sera precipitated a 60-kDa Ras oncogenes (see Milestone 17), showed that
References and links phosphorylated protein product, which they the oncogenes found in retroviruses encode
ORIGINAL RESEARCH PAPERS Nowell, P. C. The clonal consequently designated as pp60. They showed components of the normal growth-regulatory
evolution of tumor cell populations. Science 194, 23–28 (1976) | that phosphorylation of this protein was machinery of the cell.
Fearon, E. R. & Vogelstein, B. A genetic model for colorectal
tumorigenesis. Cell 61, 759–767 (1990)
essential for its function, and that it probably Many other oncogenes were shown to encode
possessed a kinase activity. This study, along proteins that bound to these kinases as upstream
with similar findings from the group of modulators or downstream signal transducers.
Raymond Erikson, indicated that protein Not only was it now apparent that tumours
oncogene research, which resulted in the iden- phosphorylation might be important in the might arise owing to the de-regulation of these
tification of more than 40 different oncogenes, transformation process. kinase-mediated signalling pathways, but the
and provided a framework for understanding In late 1979, Tony Hunter and colleagues possibility of therapeutic intervention was also
signal-transduction pathways that control identified a phosphotyrosine kinase activity in evident. The outcome of much of this early
normal cellular growth (see Milestone 16). protein immunoprecipitates of the polyoma work has been the development of therapeutic
Varmus and Bishop went on to win the Nobel virus middle T antigen. Soon thereafter, agents (see Milestone 24) that target the mutated
Prize for their discovery. studies of the Abelson murine leukaemia virus oncogenes.
Arianne Heinrichs, Chief Editor, Nature Reviews by David Baltimore and colleagues, and of RSV Ekat Kritikou, Locum Associate Editor,
Molecular Cell Biology by Hunter and Bartholomew Sefton, identified Nature Reviews Molecular Cell Biology
tyrosine-specific protein kinase activities References and links
References and links ORIGINAL RESEARCH PAPERS Levinson, A. D., Oppermann,
ORIGINAL RESEARCH PAPERS Stehelin, D., Varmus, H. E., associated with the product of the Abelson H., Levintow, L., Varmus, H. E. & Bishop, J. M. Evidence that the
Bishop, J. M. & Vogt, P. K. DNA related to the transforming transforming gene of avian sarcoma virus encodes a protein
leukaemia viral oncogene (Abl) and Src.
gene(s) of avian sarcoma viruses is present in normal avian kinase associated with a phosphoprotein. Cell 15, 561–572
DNA. Nature 260, 170–173 (1976) | Spector, D. H., Varmus, H. Both groups showed that these enzymes were (1978) | Collett, M. S. & Erikson, R. L. Protein kinase activity
E. & Bishop, J. M. Nucleotide sequences related to the essential for the malignant transformation of associated with the avian sarcoma virus src gene product.
transforming gene of avian sarcoma virus are present in DNA of Proc. Natl Acad. Sci. USA 75, 2021–2024 (1978) | Eckhart, W.,
uninfected vertebrates. Proc. Natl Acad. Sci. USA 75, 4102–
cells by the oncogenic proteins. As viral Src Hutchinson, M. A. & Hunter, T. An activity phosphorylating
was derived from an evolutionarily conserved tyrosine in polyoma T antigen immunoprecipitates. Cell 18,
4106 (1978) | Oppermann, H., Levinson, A. D., Varmus, H. E.,
925–933 (1979) | Witte, O. N., Dasgupta, A. & Baltimore, D.
Levintow, L. & Bishop, J. M. Uninfected vertebrate cells contain c-SRC cellular proto-oncogene (see Milestone Abelson murine leukaemia virus protein is phosphorylated in
a protein that is closely related to the product of the avian
15), logic dictated that all vertebrate cells vitro to form phosphotyrosine. Nature 283, 826–831 (1980) |
sarcoma virus transforming gene (src). Proc. Natl Acad. Sci. Hunter, T. & Sefton, B. M. Transforming gene product of Rous
USA 76, 1804–1808 (1979) | Collett, M. S., Erikson, E., Purchio, must contain at least one protein kinase that sarcoma virus phosphorylates tyrosine. Proc. Natl Acad. Sci.
A. F., Brugge, J. S. & Erikson, R. L. A normal cell protein similar USA 77, 1311–1315 (1980) | Ushiro, H. & Cohen, S.
phosphorylates tyrosine. Yet, the functional
in structure and function to the avian sarcoma virus transforming Identification of phosphotyrosine as a product of epidermal
gene product. Proc. Natl Acad. Sci. USA 76, 3159–3163 (1979) relationship between kinase activity and growth factor-activated protein kinase in A-431 cell membranes.
FURTHER READING Duesberg, P. H. & Vogt, P. K. Differences oncogenesis remained unclear. J. Biol. Chem. 255, 8363–8365 (1980) | Waterfield, M. D. et al.
between the ribonucleic acids of transforming and Platelet-derived growth factor is structurally related to the
nontransforming avian tumor viruses. Proc. Natl Acad. Sci. USA
Work published by the Stanley Cohen putative transforming protein p28sis of simian sarcoma virus.
67, 1673–1680 (1970) | Martin, G. S. Rous sarcoma virus: a laboratory a few months later, using cell Nature 304, 35–39 (1983) | Doolittle, R. F. et al. Simian sarcoma
function required for the maintenance of the transformed state. virus onc gene, v-sis, is derived from the gene (or genes)
membranes isolated from epidermal growth encoding a platelet-derived growth factor. Science 221,
Nature 227, 1021–1023 (1970) | Stehelin, D., Guntaka, R. V.,
Varmus, H. E. & Bishop, J. M. Purification of DNA factor (EGF)-responsive carcinoma cells, 275–277 (1983) | Downward, J. et al. Close similarity of
epidermal growth factor receptor and v-erb-B oncogene protein
complementary to nucleotide sequences required for neoplastic provided further insights. They found that sequences. Nature 307, 521–527 (1984)
transformation of fibroblasts by avian sarcoma viruses. J. Mol.
activation of the EGF receptor (EGFR)– FURTHER READING Collett, M. S. et al. A normal cell protein
Biol. 101, 349–365 (1976) | Brugge, J. S. & Erikson, R. L. similar in structure and function to the avian sarcoma virus
Identification of a transformation-specific antigen induced by an protein kinase complex resulted in protein transforming gene product. Proc. Natl Acad. Sci. USA 76,
avian sarcoma virus. Nature 269, 346–348 (1977) phosphorylation, mainly on tyrosine residues. 3159–3163 (1979)

MILESTONE 17
An important difference
transforming counterpart from the carcinoma lines. This implied that
The idea that cancer is a disease of altered genes was widely discussed the two versions of the gene were similar and any sequence difference was
among basic scientists in the 1970s. The clinching evidence that brought it subtle. Using an elegant molecular genetics strategy that has since become
to wider attention was the discovery of mutations in the genome of tumour obsolete, the Weinberg, Barbacid and Wigler groups systematically
cells that, when transferred into other cells, were sufficient to cause substituted each restriction fragment from the non-transforming allele
transformation. with the corresponding one from the transforming allele. In this way,
By the late 1970s, it was well known that retroviral oncogenes could they were able to hone in on the genetic lesion and, by the end of 1982, all
rapidly transform cells, and that the viruses had acquired these genes from three groups had discovered the same single amino-acid change: glycine
the genomes of the mammalian and avian cells that they infected (see to valine at position 12. Subsequent research has shown that this change
Milestone 15). It was therefore proposed that mutations in the cellular alters the structure of the RAS protein to make it constitutively active.
homologues of these genes could transform cells in the absence of any viral During just 1 year, not only was the concept of the cellular oncogene
involvement, and that this occurred in a substantial proportion of human confirmed by the cloning of cellular RAS, but the activating mutation
cancers. Key discoveries by the Robert Weinberg and Geoffrey Cooper was also identified. The developments of 1982 were a crucial step
groups showed that such transformation could occur when the DNA towards the modern understanding of cancer as a complex interplay
of a chemically mutagenized transformed mouse cell was transferred. between different types of genetic lesion.
However, the precise identity of the transforming gene was not known, as a Patrick Goymer, Assistant Editor,
lot of irrelevant DNA was also transferred. Nature Reviews Cancer and Nature Reviews Genetics
References and links
Finally, in 1982, the Weinberg, Michael Wigler and Mariano Barbacid ORIGINAL RESEARCH PAPERS Shih, C., Shilo, B. Z., Goldfarb, M. P., Dannenberg, A. &
groups all cloned the first oncogene, from bladder carcinoma lines, after Weinberg, R. A. Passage of phenotypes of chemically transformed cells via transfection of DNA
and chromatin. Proc. Natl Acad. Sci. USA 76, 5714–5718 (1979) | Cooper, G. M., Okenquist, S. &
closing in on the relevant DNA by numerous rounds of transfection. In Silverman, L. Transforming activity of DNA of chemically transformed and normal cells. Nature 284,
each round, more of the irrelevant DNA was lost, until the actual oncogene 418–421 (1980) | Shih, C. & Weinberg, R. A. Isolation of a transforming sequence from a human
bladder carcinoma cell line. Cell 29, 161–169 (1982) | Goldfarb, M., Shimizu, K., Perucho, M. &
could be cloned with the use of linked sequence tags. These cloned cellular Wigler, M. Isolation and preliminary characterization of a human transforming gene from T24
genes had the same transforming properties as the oncogenes from bladder carcinoma cells. Nature 296, 404–409 (1982) | Pulciani, S. et al. Oncogenes in human
retroviruses. tumor cell lines: molecular cloning of a transforming gene from human bladder carcinoma cells.
Proc. Natl Acad. Sci. USA 79, 2845–2849 (1982) | Parada, L. F., Tabin, C. J., Shih, C. & Weinberg,
Having uncovered the presence of cellular oncogenes, attention R. A. Human EJ bladder carcinoma oncogene is homologue of Harvey sarcoma virus ras gene.
turned immediately to their identity. Within a few months, the Weinberg Nature 297, 474–478 (1982) | Santos, E., Tronick, S. R., Aaronson, S. A., Pulciani, S. & Barbacid,
M. T24 human bladder carcinoma oncogene is an activated form of the normal homologue of
and Barbacid groups, as well as Cooper and colleagues, had shown BALB- and Harvey-MSV transforming genes. Nature 298, 343–347 (1982) | Der, C. J., Krontiris, T. G.
by restriction endonuclease mapping and Southern blotting that the & Cooper, G. M. Transforming genes of human bladder and lung carcinoma cell lines are
homologous to the ras genes of Harvey and Kirsten sarcoma viruses. Proc. Natl Acad. Sci. USA
oncogenes in question were the cellular homologues of the ras genes from 79, 3637–3640 (1982) | Tabin, C. J. et al. Mechanism of activation of a human oncogene. Nature
the Harvey and Kirsten sarcoma viruses. 300, 143–149 (1982) | Reddy, E. P., Reynolds, R. K., Santos, E. & Barbacid, M. A point mutation is
responsible for the activation of transforming properties by the T24 human bladder carcinoma
However, such analysis was not detailed enough to identify any oncogene. Nature 300, 149–152 (1982) | Taparowsky, E. et al. Activation of the T24 bladder
difference between the normal cellular human c-Ha-RAS1 gene and its carcinoma transforming gene is linked to a single amino acid change. Nature 300, 762–765 (1982)
MILESTONE 18
Teaming up for transformation

In the early 1980s, evidence indicated
proliferated, but then underwent crisis
that the oncogenic transformation of
and arrest. EJ RAS-expressing REFs
primary cells involved at least two stages:
were also unable to form tumours in
establishment (the immortalization of cells)
immunocompromised mice. However,
and cellular transformation. With this in
REFs expressing Myc and RAS grew rapidly
mind, Hartmut Land, Luis F. Parada and
as foci that were able to establish long-term
Robert Weinberg, working at the same time
cultures when passaged in vitro, and could
as Earl Ruley, investigated how oncogenes
form tumours in mice. These tumours were
cooperate to induce tumour development.
not metastatic, implying that beyond MYC
Weinberg and colleagues examined the
and RAS cooperation, further oncogenic
effect of expressing two recently identified
events might be required to produce an
oncogenes — an activated variant of Harvey
invasive tumour (although we now know
RAS1, EJ RAS (see Milestone 17), and either
that tumours seeded subcutaneously in
a viral or a mammalian clone of Myc — in
mice often do not metastasize). Similar
primary rat embryonic fibroblasts (REFs).
results were found by Ruley using
They found that, despite its capacity to
adenoviral E1A, polyoma virus middle T
transform rodent cell lines, EJ RAS could
antigen and T24 Harvey RAS expressed in
not transform REFs. These cells initially
baby rat kidney cells.

MILESTONE 19 inhibitor azacytidine. The reduced DNA meth-
ylation led to a decreased number of polyps in
the animals, lending support to the idea that
From obscurity to the clinic tumour-suppressor genes are hypermethylated
and silenced in cancer, and can be reactivated
by inhibiting DNA methylation.
In the early 1980s, the cancer field was abuzz activation of oncogenes, these findings implied DNA methylation inhibitors, such as
with the first discoveries of oncogenic muta- that altered DNA methylation could underlie azacytidine, are now approved for clinical use,
tions linked to cancer. The genetic mutation oncogene activation. although there is controversy about whether
responsible for the transforming properties of Later in the 1980s, the concept of tumour- they work by reactivating tumour suppressors.
the RAS oncogenes was found in 1982, to great suppressor genes, such as retinoblastoma, was Furthermore, the debate over whether altered
acclaim (see Milestone 17). In this climate, the becoming well defined (see Milestone 11). So, DNA methylation has a causal role in initiating
first observations of epigenetic abnormalities it was encouraging when relevant epigenetic cancer remains alive today. Yet, it is remarkable
in cancer were overshadowed and ignored by changes were found in these tumour-suppres- that work carried out back in 1980 by Peter
many in the field. However, studies in the 1980s sor genes. For example, Valerie Greger et al. Jones and Shirley Taylor showing the effects of
showed that epigenetic changes can occur to showed that an unmethylated CpG island at chemicals such as azacytidine on DNA meth-
both oncogenes and tumour suppressors, and the 5′ end of the retinoblastoma gene becomes ylation and cell differentiation, which attracted
have led to our present appreciation of epige- hypermethylated in tumours from retinoblas- little attention at the time, opened the door to
netic markers as diagnostics and therapeutic toma patients, leading the authors to speculate the idea of cancer treatment aimed at reversing
targets for cancer. that methylation could contribute directly to DNA methylation.
Epigenetic phenomena can be defined the silencing of tumour suppressors. Later Alex Eccleston, Senior Editor, Nature
as heritable changes in cellular information studies — such as those of Naoko Ohtani-Fujita References and links
other than the DNA sequence, which usually et al. and James Herman et al. — correlated the ORIGINAL RESEARCH PAPERS Feinberg, A. P. & Vogelstein,
B. Hypomethylation distinguishes genes of some human
involve covalent modifications to DNA or methylation of the tumour-suppressor genes cancers from their normal counterparts. Nature 301, 89–92
histones. These modifications are involved in with their actual silencing in cancer. (1983) | Greger, V., Passarge, E., Hopping, W., Messmer, E. &
controlling gene expression — for example, the More direct evidence linking DNA hyper- Horsthemke, B. Epigenetic changes may contribute to the
formation and spontaneous regression of retinoblastoma.
methylation of DNA at CpG dinucleotides in methylation with cancer formation came Hum. Genet. 83, 155–158 (1989) | Laird, P. W. et al.
gene promoters is associated with the silenc- several years later from Rudolf Jaenisch’s Suppression of intestinal neoplasia by DNA hypomethylation.
ing of transcription. In 1983, Andrew Feinberg group. They used mice carrying a ‘Min’ muta- Cell 81, 197–205 (1995)
FURTHER READING Jones, P. A. & Taylor, S. M. Cellular
and Bert Vogelstein purified DNA from several tion in the adenomatous polyposis coli (Apc) differentiation, cytidine analogs and DNA methylation. Cell 20,
primary human tumour tissues and, using gene. These mice develop intestinal polyps 85–93 (1980) | Ohtani-Fujita, N. et al. CpG methylation
inactivates the promoter activity of the human retinoblastoma
methylation-sensitive restriction enzymes, early in life and are a model system for the tumor-suppressor gene. Oncogene 8, 1063–1067 (1993) |
found lowered DNA methylation of specific early stages of human colorectal cancer. Peter Herman, J. G. et al. Silencing of the VHL tumor-suppressor
genes compared with DNA from adjacent Laird et al. reduced DNA methylation in gene by DNA methylation in renal carcinoma. Proc. Natl Acad.
Sci. USA 91, 9700–9704 (1994) | Feinberg, A. P. & Tycko, B.
normal cells. With the predominant concept Min mice by mutating a DNA methyltrans- The history of cancer epigenetics. Nature Rev. Cancer 4,
at the time being that cancer is caused by ferase gene and using the methyltransferase 143–153 (2004)
Oncogene cooperation was also studied in act in numerous ways to circumvent the pathways that have evolved to limit cellular
avian erythroid progenitor cells by Thomas RAS-induced G1 growth arrest, and RAS transformation (see Milestone 20).
Graf and colleagues. In 1986, they showed prevents MYC-induced apoptosis by Nicola McCarthy, Senior Editor,
that the non-transforming viral gene v-erbA activation of the anti-apoptotic kinase AKT. Nature Reviews Cancer
could cooperate both in vitro and in vivo B-cell lymphoma 2 (BCL2) and MYC also References and links
ORIGINAL RESEARCH PAPERS Land, H., Parada, L. F. &
with v-src, v-Ha-ras and v-erbB cooperate effectively (see Milestone 12). Weinberg, R. A. Tumorigenic conversion of primary embryo
The ability of different oncogenes BCL2 is an unusual oncogene product in fibroblasts requires at least two cooperating oncogenes. Nature
to cooperate in producing cellular that, unlike RAS, it cannot transform rodent 304, 596–602 (1983) | Ruley, H. E. Adenovirus early region 1A
enables viral and cellular transforming genes to transform primary
transformation has been a cornerstone of cell lines and, unlike MYC, it does not cells in culture. Nature 304, 602–606 (1983)
cancer research during the intervening induce proliferation. Nevertheless, in 1990, FURTHER READING Khan, P. et al. v-erbA cooperates with
years. With the development of further Andreas Strasser and colleagues showed sarcoma oncogenes in leukemic cell transformation. Cell 45,
349–356 (1986) | Strasser, A., Harris, A. W., Bathe, M. L. &
molecular techniques, it has become that BCL2 synergizes with MYC to rapidly Cory, S. Novel primitive lymphoid tumours induced in
possible to piece together how and why produce tumours in mice. This is because, transgenic mice by cooperation between myc and bcl-2.
specific oncogenes cooperate so effectively. as shown later by several groups, MYC- Nature 348, 331–333 (1990) | Bissonnette, R. P., Echeverri,
F., Mahboubi, A. & Green, D. R. Apoptotic cell death induced
Land and colleagues went on to show that induced apoptosis is suppressed by the
by c-myc is inhibited by bcl-2. Nature 359, 552–554 (1992) |
high expression levels of RAS, like those expression of BCL2, leaving the proliferative Fanidi, A., Harrington, E. H. & Evan, G. I. Cooperative
used in the experiments described above, capacity of MYC unchecked. interaction between c-myc and bcl-2 proto-oncogenes.
Nature 359, 554–556 (1992) | Wagner, A. J., Small, M. B. &
induce G1 arrest in primary cells owing to Identification of the growth-restrictive
Hay, N. Myc-mediated apoptosis is blocked by ectopic
the expression of the cell-cycle inhibitor aspects of oncogene activation, such as cell- expression of Bcl-2. Mol. Cell Biol. 13, 2432–2440 (1993) |
p21. By contrast, expression of MYC cycle arrest and apoptosis, allowed cancer Kauffman-Zeh, A. et al., Suppression of c-Myc-induced
apoptosis by Ras signalling through PI(3)K and PKB.
was found to induce both proliferation biologists to appreciate the complexities
Nature 385, 544–548 (1997) | Lloyd, A. C. et al. Cooperating
and apoptosis (see Milestone 12). MYC of the molecular cross-talk in tumour oncogenes converge to regulate cyclin/cdk complexes.
and RAS cooperate because MYC can cells, and to begin to understand the Genes Dev. 11, 663–677 (1997)

MILESTONES
MILESTONE 20
Stop or die!
Half a century ago, epidemiologists proposed
that cancers result from multiple ‘hits’
(see Milestone 9). Initially, the focus was on
dominantly acting viral oncogenes and activating
mutations in the RAS oncogene. Later, cell fusion promoter. Therefore, this paper not only uncovered upstream and
and genetic experiments showed that recessive mutations cause defects downstream events in the
in tumour suppression (see Milestone 11). Bert Vogelstein reconciled p53-dependent DNA damage-signalling pathway, but also described one
the oncogene and tumour-suppressor camps by describing how both of the first p53 target genes. The importance of these papers is threefold:
events are necessary for colorectal carcinogenesis (see Milestone 14). they explain how the cell cycle is arrested after DNA damage, and how
Arnold Levine, David Lane and colleagues discovered the first tumour- p53 loss might contribute to genetic instability and tumour formation,
suppressor gene, p53 (also known as TP53), although it was initially and they show that DNA damage elicits a signal-transduction response
described as an oncogene. Levine showed that p53 suppresses involving the gene mutated in AT (now known to be the ATM kinase),
transformation, while Vogelstein reported that both p53 alleles are p53 and p53 target genes.
mutated in colorectal cancer, a finding subsequently extended to most By the mid-1990s, it became clear that apoptosis was a key tumour-
common human tumour types, with over 20,000 p53 mutations now on suppressive pathway (see Milestone 12), and that p53 induces apoptosis
record. The second tumour suppressor to emerge was the and is required for DNA damage and oncogene-induced apoptosis.
retinoblastoma protein RB (see Milestones in Cell Division To investigate the role of p53-dependent apoptosis in brain tumour
Milestone 15). Both RB and p53 have been on the citation bestseller lists progression, Holly Symonds and colleagues used transgenic mice
ever since it became apparent that the main DNA tumour viruses expressing a SV40 T-antigen mutant that inhibits RB, but not p53.
transform cells by inactivating both RB and p53. The RB pathway is now Tumour growth relative to wild-type T antigen slowed in p53-wild type,
firmly enshrined in cell-cycle regulation, and defects in this pathway are but not in p53-null, mice; p53-heterozygous mice exhibited stochastic
a universal feature of cancer. emergence of p53-null tumours, and this correlated with decreased
In 1989, David Livingston and Ed Harlow published an early apoptosis. At the same time, Sharon Morgenbesser and colleagues
milestone: they found that RB is phosphorylated in a cell cycle- reported increased proliferation and apoptosis in the developing ocular
dependent manner, as synchronized primary and immortalized cells lens of RB-null mice; apoptosis was suppressed in RB/p53 double-null
enter the DNA-replication phase (S phase). They reported, separately, mice, indicating p53 dependence. These papers, together, are the first
that SV40 T antigen, which can drive G1-arrested cells into the cycle, to describe that inappropriate S-phase entry owing to loss of RB results
only binds unphosphorylated RB — the first indication that this is in p53-dependent apoptosis, thereby linking the two central tumour-
the growth-suppressive form of RB. Therefore, they surmised that suppressor pathways in the cell.
unphosphorylated RB acts to block exit from G1. These studies represent only a couple of the milestones in our
p53 has emerged as a crucial guardian of the genome, and several understanding of RB and p53, and their role in cell-cycle and DNA-
exceptional papers first described its role in the DNA damage- damage checkpoints, which have dominated cancer research for the
checkpoint response. It was known that both p53 and DNA damage past decade.
inhibit DNA replication, and cause G1 cell-cycle arrest. Michael Kastan Bernd Pulverer, Editor, Nature Cell Biology
and colleagues connected these findings in haematopoietic cells by References and links
ORIGINAL RESEARCH PAPERS DeCaprio, J. A. et al. The product of the retinoblastoma
showing that the G1-checkpoint arrest correlates with p53 protein
susceptibility gene has properties of a cell cycle regulatory element. Cell 58, 1085–1095 (1989) |
induction, and that this response is sensitive to caffeine — later shown Buchkovich, K., Duffy, L. A. & Harlow, E. The retinoblastoma protein is phosphorylated during
to block ATM kinase — and cycloheximide. Importantly, cells with specific phases of the cell cycle. Cell 58, 1097–1105 (1989) | Kastan, M. B., Onyekwere , O.,
Sidransky, D., Vogelstein, B. & Craig, R. W. Participation of p53 protein in the cellular response to
mutant or no p53 did not arrest in G1 after γ-irradiation (IR), while DNA damage. Cancer Res. 51, 6304–6311 (1991) | Kuerbitz, S. J., Plunkett, B. S., Walsh, W. V. &
maintaining a second checkpoint arrest in G2. In a second paper, Kastan Kastan, M. B. Wild-type p53 is a cell cycle checkpoint determinant following irradiation. Proc. Natl
generalized these findings and showed that re-expression of p53 in Acad. Sci. USA 89, 7491–7495 (1992) | Kastan, M. B. et al. A mammalian cell cycle checkpoint
pathway utilizing p53 and GADD45 is defective in ataxia-telangiectasia. Cell 71, 587–597 (1992) |
p53-null cells rescued the IR-induced G1-checkpoint arrest. Conversely, Symonds, H. et al. p53-dependent apoptosis suppresses tumor growth and progression in vivo.
a p53 mutant was able to abrogate the G1 checkpoint in p53 wild-type Cell 78, 703–711 (1994) | Morgenbesser, S. D., Williams, B. O., Jacks, T. & DePinho, R. A.
cells in a dominant–negative fashion. A third paper by Kastan placed p53-dependent apoptosis produced by Rb-deficiency in the developing mouse lens. Nature 371,
72–74 (1994)
p53 in a checkpoint-signalling pathway; he noted that cells from ataxia FURTHER READING DeLeo, A. B. et al. Detection of a transformation-related antigen in
telangiectasia (AT) patients also lacked the G1 DNA-damage checkpoint chemically induced sarcomas and other transformed cells of the mouse. Proc. Natl Acad. Sci. USA
and, proposing that the defects in AT and p53 are functionally linked, he 76, 2420–2424 (1979) | Lane, D. P. et al. T antigen is bound to a host protein in SV40-transformed
cells. Nature 278, 261–263 (1979) | Linzer, D. I. et al. Characterization of a 54K dalton cellular SV40
documented a decreased p53 induction in AT cells after IR. tumor antigen present in SV40-transformed cells and uninfected embryonal carcinoma cells. Cell
Importantly, this paper used primary embryonic fibroblasts from 17, 43–52 (1979) | Baker, S. J. et al. Chromosome 17 deletions and p53 gene mutations in
p53-null mice, rather than transformed cell lines. Just previously, colorectal carcinomas. Science 244, 217–221 (1989) | Finlay, C. A. et al. The p53 proto-oncogene
can act as a suppressor of transformation. Cell 57, 1083–1093 (1989) | Fields, S. et al. Presence of
p53 had been shown to be a sequence-specific DNA-binding protein a potent transcription activating sequence in the p53 protein. Science 249, 1046–1049 (1990) |
capable of transcriptional activation. Furthermore, it was known that Raycroft, L. et al. Transcriptional activation by wild-type but not transforming mutants of the p53
the radiation sensitive GADD45 gene was not induced in AT and several anti-oncogene. Science 249, 1049–1051 (1990) | Lane, D. P. Cancer. p53, guardian of the
genome. Nature 358, 15–16 (1992) | Marte, B. Oncogenes and the retinoblastoma pathway.
tumour cell lines. Kastan showed that GADD45 induction requires Milestone 15. Nature Rev. Mol. Cell Biol. 2, S12 (2001) | Sherr, C. et al. The RB and p53 pathways
p53, and that wild-type p53 bound to a p53 consensus site in the gene in cancer. Cancer Cell 2, 103–112 (2002)

MILESTONE 21 A breakthrough came with the identification of
genes on chromosomes 2 and 17 that were asso-
ciated with major fractions of colon and breast
The human touch cancers, respectively. These, and related genes
discovered shortly thereafter, were MSH2 and
MLH1 in hereditary non-polyposis colorectal
Although we note that cancer is a complex cancer, and BRCA1 and BRCA2 in hereditary
and challenging disease that has been attacked breast cancer syndromes. Interestingly, in both Internet has greatly assisted this process, and
on many research fronts, we must also be types of hereditary cancer, as well as in other websites such as that of the National Cancer
aware that there is a human side to dealing cancer-predisposition disorders, the predispos- Institute in the United States are consulted by
with cancer. ing genes affect DNA repair rather than cell thousands of patients and their family members
In the late 1980s and early 1990s, the genetic growth per se (see Milestone 22). each day.
basis for cancer-predisposition syndromes rap- The penetrance of mutations varies in fami- Chris Gunter, Senior Editor, Nature
idly unravelled with the discovery of a number lies, and most breast and colon cancers do not References and links
of tumour-suppressor genes that were found appear to be of hereditary origin. Yet, for fami- ORIGINAL RESEARCH PAPERS Call, K. M. et al. Isolation and
characterization of a zinc finger polypeptide gene at the human
to be inherited in mutant form in affected lies dealing with these cancers, the identification chromosome 11 Wilms’ tumor locus. Cell 60, 509–520 (1990) |
families. These included genes associated of the genes meant that testing was possible, and Cawthon, R. M. et al. A major segment of the neurofibromatosis
type 1 gene: cDNA sequence, genomic structure, and point
with the childhood cancer Wilms’ tumour, Li difficult prophylactic-care decisions could be mutations. Cell 62, 193–201 (1990) | Gessler, M. et al.
Fraumeni syndrome, neurofibromatosis, famil- informed by the test results. Homozygous deletion in Wilms tumours of a zinc-finger
gene identified by chromosome jumping. Nature 343, 774–778
ial adenomatous polyposis, von Hippel-Landau In the decade since their discovery, testing (1990) | Hall, J. M. et al. Linkage of early-onset familial breast
disease and retinoblastoma (see Milestone 11). for cancers resulting from mutations in cancer- cancer to chromosome 17q21. Science 250, 1684–1689 (1990) |
Malkin, D. et al. Germ line p53 mutations in a familial syndrome
Many of these genes are now known to have key susceptibility genes has become more common. of breast cancer, sarcomas, and other neoplasms. Science 250,
roles in cell proliferation, cell-cycle checkpoints However, the decision to be tested and knowing 1233–1238 (1990) | Srivastava, S. et al. Germ-line transmission
of a mutated p53 gene in a cancer-prone family with Li-Fraumeni
and cell death (see Milestone 20). what to do with the information have not neces- syndrome. Nature 348, 747–749 (1990) | Wallace, M. R. et al.
At the same time, researchers were franti- sarily become any easier. For example, prenatal Type 1 neurofibromatosis gene: identification of a large transcript
disrupted in three NF1 patients. Science 249, 181–186 (1990) |
cally searching for, and finding, more suscepti- testing is now available for conditions such as Nishisho, I. et al. Mutations of chromosome 5q21 genes in FAP
bility genes and molecular mechanisms, so that neurofibromatosis, which is a common heredi- and colorectal cancer patients. Science 253, 665–669 (1991) |
Groden, J. et al. Identification and characterization of the familial
genetic counsellors and patient-care profes- tary disease that leads to numerous benign adenomatous polyposis coli gene. Cell 66, 589–600 (1991) |
sionals around the world could begin translat- tumours throughout the body. Fishel, R. et al. The human mutator gene homolog MSH2 and its
association with hereditary nonpolyposis colon cancer. Cell 75,
ing these important advances into real-world Unfortunately, like MSH2 and BRCA1, the 1027–1038 (1993) | Leach, F. S. et al. Mutations of a mutS
advice for patients who were forced to make neurofibromatosis gene NF1 is large and subject homolog in hereditary nonpolyposis colorectal cancer. Cell 75,
profound decisions. The discovery of inherited to a range of mutations, so testing is difficult 1215–1225 (1993) | Peltomaki, P. et al. Genetic mapping of a
locus predisposing to human colorectal cancer. Science 260,
mutations leading to more common diseases, unless a specific mutation has been identified 810–812 (1993) | Bronner, C. E. et al. Mutation in the DNA
such as breast cancer and colon cancer, serves in another family member. Moreover, as in mismatch repair gene homologue hMLH1 is associated with
hereditary non-polyposis colon cancer. Nature 368, 258–261
to illustrate this point. hereditary breast and colon cancer syndromes, (1994) | Futreal, P. A. et al. BRCA1 mutations in primary breast
Before 1990, we knew that 5–10% of the chance of developing tumours and their and ovarian carcinomas. Science 266, 120–122 (1994) |
Miki, Y. et al. A strong candidate for the breast and ovarian
breast and colon cancers occurred in familial severity varies greatly, even when a mutation is cancer susceptibility gene BRCA1. Science 266, 66–71 (1994) |
patterns, but whether these were owing to found. Basic cancer research must therefore lead Papadopoulos, N. et al. Mutation of a mutL homolog in
hereditary colon cancer. Science 263, 1625–1629 (1994) |
shared environments, several interacting directly to the translation of important findings Wooster, R. et al. Identification of the breast cancer susceptibility
genes or single major genes was not known. into information for patients. The advent of the gene BRCA2. Nature 378, 789–792 (1995)
MILESTONE 22 When cells are exposed to ultraviolet (UV) light, base adducts are
formed that must be excised for replication to occur. This process,
nucleotide-excision repair (NER), involves recognition of distortion
Indirect but just as effective of the DNA helix, assembly of a complex on and around the lesion,
and excision of a single-strand fragment containing the modified
base. Several human syndromes show UV hypersensitivity, and one,
Although many cancers result from mutations in prototype onco- xeroderma pigmentosum (XP), displays a strong predisposition to
genes and tumour-suppressor genes that regulate cell proliferation skin cancer.
and apoptosis (see Milestones 10, 11, 12, 20 and 21), cancer can XP patients were originally classified in eight complementation
also arise indirectly from defects in the protective cellular mecha- groups. In 1990, two human genes with roles in NER were cloned,
nisms that repair DNA damage. This idea originated in the study by and both were linked to XP. The sequence of excision-repair cross-
Theodor Boveri of chromosomal imbalances in somatic cells (see complementing 3 (ERCC3; also known as XPB), cloned by Geert Weeda
Milestone 2). The type of DNA damage can range from the subtle, et al., implied that it encoded a DNA helicase. Complementation
such as a single unrepaired base lesion, through small deletions or studies showed that in the unique XP group B individual, a splic-
insertions, to macroscopic changes that manifest as non-reciprocal ing mutation in ERCC3 resulted in a frameshift. Kiyoji Tanaka et
chromosome translocations (see Milestone 10). Genetic instability al. later cloned the XP group A-complementing protein (XPA; also
at any of these levels can predispose to cancer by increasing the known as XPAC) gene, the mRNA of which was reduced in cells from
rate at which potentially oncogenic mutations and chromosomal XP-A individuals. From its sequence, XPA was proposed to promote
▼
alterations occur. incision surrounding the lesion.

MILESTONES
The association between XP and DNA repair deficiency arose

▼
MILESTONE 23
from the extreme UV sensitivity of the patients, rather than specific
observations of damage at the DNA level. In patients with hereditary
non-polyposis colon cancer (HNPCC), however, the link to defective
repair was obvious: microsatellite repeat sequences in their cells had
Profiling cancer expression
changes similar to those found in bacterial mismatch repair (MMR)- Tailoring cancer therapy to specific tumour types maximizes efficacy
deficient mutants. while minimizing toxicity. Historically, cancer classification has
This observation encouraged efforts to locate human genes with been based on morphology, but cancers with seemingly identical
homology to the bacterial MMR proteins MutS, MutH and MutL. In morphological and histopathological features can progress and respond
1993, two groups using complementary approaches identified MutS to therapy in radically different ways. A better method of classifying
homologue 2 (MSH2) as an HNPCC-associated gene. Whereas Richard cancers was needed to help predict clinical outcome and make the most
Fishel et al. went directly after homologues of MutS using a degenerate of the available therapy — the possible solution came from microarray
primer strategy, Frederick Leach et al. used markers linked to HNPCC technology.
to define the disease locus, and then isolated the candidate MMR gene. The first evidence that gene-expression profiling could distinguish
Additionally, Leach et al. reported that chromosome 2-linked HNPCC between cancer types came in 1999, from Todd Golub, Donna Slonim
families had mutations in MSH2. A few months later, in 1994, the and colleagues. They chose two types of leukaemia as a test case: acute
myeloid and acute lymphoblastic. The approach involved identifying
gene responsible for chromosome 3-linked HNPCC was cloned by
a ‘predictor class’ of genes, based on their non-random expression
Nickolas Papadopoulos et al. and Eric Bronner et al. Not surprisingly,
patterns, and evaluating the prediction strength. In addition to
this turned out to be the human MutL homologue, MLH1.
distinguishing between the two types of leukaemia on the basis of
Gross chromosomal changes are consistently observed in human
expression-profile differences, the method could also predict their
cancers, and their mechanistic basis is the subject of active investiga- responsiveness to chemotherapy. The paper laid out a general analytical
tion. One line of research indicates that the combination of telomerase approach to cancer classification based on gene expression, which could
dysfunction and p53 inactivation leads to chromosome instability. Late- be adapted to assign cancers to hitherto unknown classes.
passage telomerase-deficient mice were known to have shortened telom- A year later, Ash Alizadeh, Michael Eisen and colleagues used a
eres and chromosome instability, but cell viability was compromised. By similar approach to uncover gene-expression heterogeneity in diffuse
introducing p53 deficiency into this background, Ronald DePinho and
colleagues were able to show that cell survival could be promoted, allow-
ing neoplastic transformation to occur. Furthermore, Steven Artandi et MILESTONE 24
al. found that in telomerase- and p53-deficient epithelial cells, telom-
eres become progressively shortened, leading to a rise in chromosomal
instability (non-reciprocal translocations and end-to-end fusions) and
accelerated carcinogenesis. Another line of research implies that the
Precision weapons
maintenance of the mitotic-spindle checkpoint is essential for chromo-
some stability in cancer cells. Sandra Hanks et al. found that mutations The phrase ‘the war against cancer’ might
of the spindle-checkpoint gene BUB1B caused a cancer-predisposition have become clichéd over the decades, but it
syndrome characterized by premature chromosome separation. Other does help to portray how much we have
cancer-predisposition syndromes caused by alterations in genes asso- relied on advances in weaponry to score
ciated with chromosome-level repair are ataxia telangiectasia, Bloom numerous victories against the disease.
syndrome and hereditary breast cancers (see Milestone 21). Tamoxifen proved that cancer treatments can
These and other studies established that DNA repair defects of vari- behave like ‘magic bullets’ (see Milestone 5)
ous forms and severity initiate genetic instability that affects cancer and avoid the toxic effects of traditional
development. Whether genetic instability is actually mandatory for chemotherapy treatments. Yet, the discovery
cancer development in non-familial cancers remains controversial,
although these findings stress the importance of protecting the integ-
rity of the genome as a tumour-suppression mechanism.
Angela K. Eggleston, Senior Editor, Nature
ORIGINAL RESEARCH PAPERS Weeda, G. et al. A presumed DNA helicase encoded by
ERCC-3 is involved in the human repair disorders xeroderma pigmentosum and Cockayne’s
syndrome. Cell 62, 777–791 (1990) | Tanaka, K. et al. Analysis of a human DNA excision repair
gene involved in group A xeroderma pigmentosum and containing a zinc-finger domain. Nature
348, 73–76 (1990) | Fishel, R. et al. The human mutator gene homolog MSH2 and its association
with hereditary nonpolyposis colon cancer. Cell 75, 1027–1038 (1993) | Leach, F. S. et al.
Mutations of a mutS homolog in hereditary nonpolyposis colorectal cancer. Cell 75, 1215–1225
(1993) | Papadopoulos, N. et al. Mutation of a mutL homolog in hereditary colon cancer. Science
263, 1625–1629 (1994) | Bronner, C. E. et al. Mutation in the DNA mismatch repair gene
homolog hMLH1 is associated with hereditary non-polyposis colon cancer. Nature 368, 258–
261 (1994) | Chin, L. et al. p53 deficiency rescues the adverse effects of telomere loss and
cooperates with telomere dysfunction to accelerate carcinogenesis. Cell 97, 527–538 (1999) |
Artandi, S. E. et al. Telomere dysfunction promotes non-reciprocal translocations and epithelial
cancers in mice. Nature 406, 641–645 (2000) | Cahill, D. P. et al. Mutations of mitotic checkpoint
genes in human cancers. Nature 392, 300–303 (1998) | Hanks, S. et al. Constitutional
aneuploidy and cancer predisposition caused by biallelic mutations in BUB1B. Nature Genet. 36,
1159–1161 (2004) | Ellis, N. A. et al. The Bloom’s syndrome gene product is homologous to
RecQ helicases. Cell 83, 655–666 (1995) | Savitsky, K. et al. A single ataxia telangiectasia gene
with a product similar to PI-3 kinase. Science 268, 1749–1753 (1995)

MILESTONES
large B-cell lymphoma (DLBCL), which is the most common type

of non-Hodgkin’s lymphoma. The expression profiles identified two
distinct forms of DLBCL and correlated with the responsiveness of the
tumours to treatment.
The next landmark example of how expression profiling can help
to predict clinical outcomes came from the breast cancer field. In
this case, specific molecular signatures (of genes involved in the cell
cycle, invasion, metastasis and angiogenesis) were shown to accurately
predict high likelihood of metastases and, therefore, poor overall
prognosis, in the absence of other indicators. This study was the
first to show that metastatic potential can be gleaned from the gene-
expression data of the primary tumours. After further refinement,
related breast cancer-profiling diagnostics have since become
commercially available.
Although it might still be too early to see the effect of this technology
in the clinic, an important feature of microarray analysis is its lack
of bias, which allows microarray-based cancer classification to be
systematic and not limited by our previous biological knowledge.
Magdalena Skipper, Chief Editor,
Nature Reviews Genetics
ORIGINAL RESEARCH PAPERS Golub, T. R. & Slonim, D. K. et al. Molecular classification of
cancer: class discovery and class prediction by gene expression monitoring. Science 286,
531–537 (1999) | Alizadeh, A. A. & Eisen, M. B. et al. Distinct types of diffuse large B-cell
lymphoma identified by gene expression profiling. Nature 403, 503—511 (2000) | van’t Veer, L. J.,
Dal, H. & van de Vijver, M. J. et al. Gene expression profiling predicts clinical outcome of breast
cancer. Nature 415, 530—526 (2002)
of oncogenes (see Milestones 4, 15 and 17) to grow and proliferate continuously (see will develop resistance, and Charles Sawyers
offered the possibility of creating ‘laser- Milestone 10). Imatinib mesylate was and colleagues showed that this is also
guided’ treatments — drugs that strike at the rationally designed to block the BCR–ABL determined by the target protein. Six out of
heart of tumours by zeroing in on the genetic active site, and when Brian Druker and the nine patients studied, who had relapsed
abnormalities that make cells grow colleagues carried out the first trial with after imatinib mesylate treatment, acquired
uncontrollably. the drug they found that almost all patients the same amino-acid substitution in the ABL
The first of these molecular-targeted (98%) with therapy-resistant CML saw their kinase domain, which affects the interaction
treatments was a monoclonal antibody blood counts return to normal. Yet, it turns of the drug with the kinase; the other three
called trastuzumab (Herceptin; Genentech). out that imatinib mesylate is not as selective showed BCR–ABL gene amplification.
Trastuzumab blocks the human epidermal as first thought, and this promiscuity could Understanding the molecular
growth factor receptor 2 (HER2) protein that help to treat other cancers. George Demetri underpinnings of response and resistance
is overexpressed by gene amplification in and colleagues were the first to show that to these, and other molecular-targeted
around 25% of breast cancer cases. Patients imatinib mesylate could treat patients with treatments, is helping to create a new wave of
with this form of breast cancer have a worse advanced gastrointestinal stromal tumours drugs that can harness or circumvent these
prognosis; however, in the first trial carried by blocking c-KIT. mechanisms — some of which are already
out with trastuzumab, Dennis Slamon and Designing targeted drugs for more beginning to enter the clinic. The war
colleagues found that women with advanced common and complex cancers, however, against cancer might be far from being won,
breast cancer who received the new drug as presents added challenges, as illustrated by but the era of molecular-targeted treatments
well as the usual chemotherapy fared better the story of gefitinib (Iressa; AstraZeneca). could prove to be one of the most important
than those who received chemotherapy Gefitinib blocks the activity of a tyrosine turning points in determining the outcome.
alone. kinase called epidermal growth factor Simon Frantz, News Editor,
If trastuzumab proved that molecular- receptor (EGFR) that is overexpressed in Nature Reviews Drug Discovery
targeted treatments could effectively treat 40–80% of lung cancers. Yet, gefitinib turns References and links
cancer, then a drug for chronic myeloid out to be effective in only 10–19% of lung ORIGINAL RESEARCH PAPERS Slamon, D. J. et al. Use of
chemotherapy plus a monoclonal antibody against HER2 for
leukaemia (CML) called imatinib mesylate cancer patients. Thomas Lynch, Daniel metastatic breast cancer that overexpresses HER2.
(Glivec; Novartis) changed our thinking Haber and colleagues explained how the N. Engl. J. Med. 344, 783–792 (2001) | Druker, B. J. et al.
about the power of designing such therapies. target protein governs whether the drug will Efficacy and safety of a specific inhibitor of the BCR–ABL
tyrosine kinase in chronic myeloid leukemia. N. Engl. J. Med.
CML is a rare cancer that is characterized work. Patients who respond to gefitinib have 344, 1031–1037 (2001) | Demetri, G. D. et al. Efficacy
by the union of chromosomes 9 and 22, specific mutations clustered around the ATP- and safety of imatinib mesylate in advanced gastrointestinal
stromal tumors. N. Engl. J. Med. 347, 472–480 (2002) | Lynch,
which fuses two genes called breakpoint binding pocket of the EGFR protein where
T. J. et al. Activating mutations in the epidermal growth factor
cluster region (BCR) and Abelson murine the drug binds, whereas patients who do not receptor underlying responsiveness of non-small-cell lung
leukaemia viral oncogene homologue 1 respond tend not to carry these mutations. cancer to gefitinib. N. Engl. J. Med. 350, 2129–2139 (2004) |
Gorre, M. E. et al. Clinical resistance to STI-571 cancer therapy
(ABL; also known as ABL1), to form a Equally important as knowing who will caused by BCR–ABL gene mutation or amplification. Science
tyrosine kinase that signals myeloid cells respond to treatments is knowing who 293, 876–880 (2001)


Milestones in Cancer NATURE 0

Uploaded by

Copyright:

Available Formats

Milestones in Cancer NATURE 0

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Milestones in Cancer NATURE 0

Uploaded by

Copyright:

Available Formats

MILESTONES

Observations from a ploughman

NATURE MILESTONES | C ANCER APRIL 2006 | S7

MILESTONE 2 which he had also discovered and named. By

Hide and seek colleagues, that naturally arising tumours were

S8 | APRIL 2006 www.nature.com/milestones/cancer

NATURE MILESTONES | C ANCER APRIL 2006 | S9

S10 | APRIL 2006 www.nature.com/milestones/cancer

NATURE MILESTONES | C ANCER APRIL 2006 | S11

S12 | APRIL 2006 www.nature.com/milestones/cancer

leukaemia cells. She went on to discover Cloning of the breakpoints of other

NATURE MILESTONES | C ANCER APRIL 2006 | S13

The road less

S14 | APRIL 2006 www.nature.com/milestones/cancer

NATURE MILESTONES | C ANCER APRIL 2006 | S15

MILESTONE 14 Milestone 19), he suggested that, as the result of

MILESTONE 15 chicken cells, and to less-conserved sequences

S16 | APRIL 2006 www.nature.com/milestones/cancer

NATURE MILESTONES | C ANCER APRIL 2006 | S17

Teaming up for transformation

S18 | APRIL 2006 www.nature.com/milestones/cancer

NATURE MILESTONES | C ANCER APRIL 2006 | S19

S20 | APRIL 2006 www.nature.com/milestones/cancer

alterations occur. incision surrounding the lesion.

NATURE MILESTONES | C ANCER APRIL 2006 | S21

The association between XP and DNA repair deficiency arose

S22 | APRIL 2006 www.nature.com/milestones/cancer

large B-cell lymphoma (DLBCL), which is the most common type

NATURE MILESTONES | C ANCER APRIL 2006 | S23

You might also like