THURSDAY, 6 MAY 2021Cells can be seen as machines driven by biological components and cogs, which work in harmony to allow for cell survival, growth, and death. In cancer, these finely-tuned and elegantly orchestrated processes are disrupted by one or more broken cogs. The result? Cells proliferate carelessly, invade healthy body parts, and ultimately cause irreparable damage. It comes as no surprise that precisely these mechanisms by which cells disobey programmed cell fate and turn into ever-replicating cancer cells have been the protagonists of countless research projects and scientists’ investigations. 50 years ago, in 1970, G.S. Martin and colleagues at the University of California Berkeley revealed a crucial mechanism by which specific genes termed oncogenes can be responsible for the onset of cancer. They were the first to identify some of the broken cogs that cause cells to lose track of their normal function. V-SRC was the first ever confirmed oncogene and it came from a chicken retrovirus causing sarcomas. This discovery represented an incredible stepping stone for cancer treatment by offering new druggable targets and diagnostic markers. 50 years later, our understanding of oncogenes and their role in cancer has dramatically progressed, enabling more targeted and efficacious therapies. We celebrate this 50-year-old discovery by retracing its history and the impact it has had on our understanding of the mechanisms behind cancer onset.
A chicken virus can transform healthy cells into cancer cells | The history of the first oncogene is rooted in observations made at the start of the 20th century when P. Rous, a scientist at the Rockefeller Institute in New York, was presented with a hen bearing a large tumour. He found that the hen’s cancer could be transplanted and propagated in chickens. Rous went further and showed that the tumour could be transmitted using cell-free filtrates. He concluded there was a ‘filtrable agent’ later called Rous Sarcoma virus (RSV) that caused tumour onset in chickens. His discovery was met by the scientific community with skepticism. Most experts thought that transmissible cancers pertained only to animals and held no value in human cancer research. Rous even reported being told ‘My dear fellow, don’t you see, this can’t be cancer because you know its cause’. With time, more light was shed on viruses and their properties, such as the fact they carry genetic information in the form of RNA or DNA and use information encoded in their genes to infect and replicate. Viral genomes became a hot topic of research and various experiments were designed to help disentangle and identify the function of various viral genes. These settings primed the field for the discovery of the first oncogene.
In fact, Rous’s discovery never fell into oblivion as labs continued researching the RSV virus. Some found that certain tumours induced by RSV did not appear to produce more viral particles once infected. This observation allowed scientists to isolate strains of virus that could infect cells and transform them into cancer cells, but lacked the ability to cause viral replication: these strains were called replication-defective. A few years later, other RSV strains that had opposite characteristics were identified. These strains caused viral replication, but infected cells were not transformed into cancer cells. Taken together, these findings showed that replication and transforming abilities were genetically different and separable characteristics carried by the RSV genome.
Finding the first oncogene, V-SRC | Researchers worked ceaselessly to separate replication-defective strains from transformation-defective strains. At the same time, other scientists identified RSV variants that induced different phenotypes in the cells they transformed. These mutant variants induced the formation of spindle-shaped cells, while infection with common RSV caused cells to round up. The characterisation of RSV strains that induced different morphologies in transformed cells suggested that the phenotype of the cancer cell is controlled by the genetic information carried by the viral genome. H. Temin, the scientist who published this result concluded: ‘Mutational differences in the virus determine the various morphologies of the infected cells, thus the virus becomes equivalent to a cellular gene controlling cell morphology’ moving a step closer to the identification of the gene behind the transformation into cancer. The year 1970 marks a turning point in the first oncogene history. In that year, G.S. Martin at UC Berkeley isolated a temperature sensitive strain that at the nonpermissive temperature failed to transform cells but continued to replicate. Moreover, when cells transformed by this mutant strain were shifted to the nonpermissive temperature, they reverted to their normal appearance losing the cancer-like phenotype, indicating that the mutant function was needed to maintain the transformed cancer-like state. This unequivocally indicated the existence of a viral gene that is necessary for cell transformation but dispensable for replication. Shortly after, the incredible work by two other scientists led to the physical identification of the gene. P.H. Duesberg and P.K. Vogt isolated the genome of non-transforming strains and of transforming strains. The latter always contained a slightly bigger RNA subunit called ‘a’ while non-transforming strains contained a shorter subunit ‘b’. They concluded that the genetic material missing from the b subunit was the gene responsible for transformation of cells i.e. an oncogene. Hence, the first oncogene V-SRC was physically identified.
From viruses to humans, C-SRC is the first human proto-oncogene | The following step in the history of V-SRC had a disruptive effect in the fields of tumour virology and cancer genetics. M. Bishop and H. Varmus at UC San Francisco were surprised that RSV didn’t need V-SRC for replication and wondered why the virus would carry a seemingly unnecessary gene. Could it be that it was acquired from cells? Using viral genetic material as a template, Bishop and Varmus made a probe to see if it would bind and recognise similar genetic sequences in the DNA of birds. This was indeed the case and it was found that several species contained sequences closely related to V-SRC. The animal version of the gene was called C-SRC. It became clear that viruses could acquire this gene during their replication cycle. Free from careful cellular regulation, the V-SRC gene becomes a broken cog constitutively expressing active protein and driving the cell in continuous growth and division.
It was found that whenever overexpressed or mutated, C-SRC caused cells to transform in a cancer-like state in a similar fashion to the RSV-mediated cell transformation. The term ‘proto-oncogene’ was coined to describe genes that were not intrinsically oncogenic but could become so if mutated or over-expressed. Bishop and Varmus’ landmark discovery of C-SRC stimulated a burst of research into proto-oncogenes. Spearheading this new approach to cancer research, Vogt and Duesberg studied other viruses that cause avian blood cancers and identified other proto-oncogenes, additional cogs that could be broken or modified leading to the transformation of healthy cells into cancer cells. These novel oncogenes were later also shown to be derived from cellular oncogenes known today as some of the most important drivers of human cancer.
MYC, RAS, ERBB: the trio that consolidated the view of cancer as genetic disease | One of the first oncogenes to be identified was ERBB. This oncogene of avian retroviruses can induce an acute form of erythroid leukemia called erythroblastosis and biochemistry studies revealed that the viral protein ERBB protein was closely related to an important family of human proteins. In 1984, J. Downward and colleagues reported its sequence similarity to a human gene EGFR setting the scene for several studies involving the human ERBB/EFGR gene and its role in cancer. Various studies proved that EGFR can function as an oncogenic ‘driver’ of cancer in diverse human tumours. For instance, mutations that mechanistically resemble those seen in the viral oncogene occur in the EGFR of glioblastoma multiforme and non-small cell lung cancer. Genetic studies then showed that the human genome contains three additional genes that are closely related to EGFR: HER2, HER3, and HER4. HER2 is frequently expressed at high levels in breast cancer and its oncogenic potential became clear when DNA from tumours containing amplified HER2 was shown to turn healthy cells into cancer cells.
MYC was another one of the first oncogenes that emerged after SRC. The function of the MYC protein was then investigated and a fundamental insight was offered by the observation that MYC would localise in the nucleus of cells. The immediate reaction was to think it could regulate DNA expression, though investigating the protein’s role was easier said than done. It took substantial effort to find that MYC proteins must couple to MAX proteins in order to bind DNA. Following this realisation, scientists revealed a myriad of cellular pathways affected by the activity of MYC. Indeed, cellular levels of MYC are tightly regulated, and when MYC is overexpressed cells can undergo either uncontrolled replication or apoptosis, depending on a multitude of factors that are still not completely elucidated. Recent studies have shown that in cells overexpressing MYC, the MYC protein can bind to more than 7,000 genomic locations (in jargon, loci) demonstrating MYC’S potential for disruptive regulation of cellular activities. Furthermore, it has become clear the role of MYC goes beyond a proto-oncogene that can drive cancer formation. In fact, MYC can also affect resistance to cancer treatments and can cause regression in certain types of cancers driven by other mutated oncogenes.
One such gene is RAS, which also emerged from research on viruses. A series of pivotal experiments linked RAS directly to human cancer. Initially, transfer of DNA from human cancer cells was found to transform mouse cells. The eureka moment came with the discovery that the transforming DNA derived from human cancer cells was homologous to the RAS oncogene isolated from Harvey sarcoma virus and Kirsten sarcoma virus. Within two years, 1982 to 1984, the findings of c-MYC in Burkitt lymphoma, N-MYC in neuroblastoma and oncogenic RAS in diverse human cancers, linked these oncogenes to human tumours. The discoveries with MYC, RAS, and ERBB have special historical significance, because they consolidated the view of cancer as a genetic disease. The significance of identifying oncogenes in viruses was at first exclusively theoretical and experimental, but it then showed that normal vertebrate cells could be transformed into cancer cells by the action of a single gene. This was the revolutionary insight offered by all the scientists who dedicated their lives to the very early stages of cancer genetics research.
Oncogene addiction and targeted therapy | We now know that MYC is involved in several cancer types including breast, colorectal, pancreatic, gastric, and uterine cancers. The ERBB/EGFR gene family also represents an important proto-oncogene group: EGFR overexpression is often associated with many cancers such as gliomas and non-small-cell lung carcinoma, while ERBB-2 overexpression can occur in breast, ovarian, bladder, non-small-cell lung carcinoma, as well as several other cancer types. Mutations in the RAS family of proto-oncogenes are very common, occurring in an estimated 20% to 30% of all human tumours. Research on the trio of human cancer (RAS, MYC, and ERBB) fostered further research into human oncogenes. If we define an oncogene as a replication-promoting gene whose product is overly active in cancer, then the number of oncogenes we now know is probably in excess of a thousand and growing. The arsenal of oncogenic genetic alterations would not be complete without the multitude of mutations that cause the loss of function of genes that act as tumour suppressors and the role of microRNAs, which are small non-coding RNA molecules that can have anti- or pro-oncogenic activity. However, this complex picture fades in importance when some cancers show a striking dependency on one single oncogene. The growth and survival of oncogene-dependent cancers can often be impaired by the inactivation of a single oncogene. This so-called oncogene addiction represents the ideal basis for creating targeted therapies against cancer.
Finding what oncogene cancers are addicted to is equivalent to finding Achilles’ heel. If a drug is then used to target the right oncogene or switch it off, cancers succumb. For instance, switching on the C-MYC oncogene in blood cells leads to the development of leukemias in a transgenic mouse model. When this gene was switched off, the leukemia cells stopped proliferating and several died. Evidence supporting the concept of oncogene addiction is strong and has guided important advancements in our therapeutic approaches to cancer. For example, antibodies have been designed to target specific oncogenes in human cancers and their efficacy probably represents some of the most convincing evidence for the importance of oncogene addiction. One of the earliest examples is the antibody trastuzumab (Herceptin), which targets the receptor protein HER-2/NEU in breast cancer. Herceptin was approved at the end of the 90s in the US and arrived in the EU in the year 2000. Its efficacy looked so promising from the very first trials that the drug was fast-tracked by regulators (FDA) so that it could promptly reach the clinic. It is now on the World Health Organization's List of Essential Medicines, a list reserved to the medications considered to be most effective and safe to meet the most important needs in a health system. Other success stories of targeted therapy are provided by drugs such as imatinib, which targets the BCR-ABL oncogene product in chronic myeloid leukemia and gefitinib and erlotinib, which target EGFR in non-small cell lung carcinoma, pancreatic cancer, and glioblastoma.
Although our understanding of oncogene addiction has led to disruptive results in cancer treatment and scientific evidence does indeed show that cancers can depend on a single oncogene at certain points in time, it is apparent from mouse models and clinical experience with targeted drugs that cancers can ‘evade’ oncogene addiction. This can be due to mutations in other genes and cellular pathways that ensure survival and uncontrolled growth, thus escaping targeted therapy aimed at hitting the cancer in its Achilles’ heel. For this reason, treatment resistance can arise and it is unlikely that the use of a single targeted therapy will achieve long-lasting remissions or cures in human cancers. This is especially true for late-stage cancers which often present high mutational loads and heterogeneity, giving the cancer many ways to escape from one specific type of oncogene addiction.
Thus, combining different types of therapy becomes necessary: this approach is called combination therapy.
Combination therapy: biochemistry & physics cocktails | Clinical studies have indicated that targeted therapy can be enhanced by combination with cytotoxic agents which act by inhibiting and interfering with DNA or chromosomal replication. The design of combination therapies has also gone beyond a mere biochemical approach of pairing two or more drugs by leveraging technologies such as radiotherapy in combination with pharmaceutical agents. Radiotherapy uses X-rays and gamma-rays, (high-energy radiation in the electromagnetic spectrum), to ionise atoms, causing the loss of an electron and leaving a positively charged ion. Ionised atoms, crucially, can disrupt the functions of various biological molecules in cells and ultimately cause cell death. Beams of radiation are produced by machines called linear accelerators that cause high-speed electrons (other particles may be used) to collide with a metal target, releasing photons. The radiation waves in the form of photon beams can then be targeted to the patient’s cancer. At the cellular level, radiotherapy mainly causes indirect DNA damage. Radiation causes the formation of free radicals which are toxic to cells and cause breaks in the cell’s DNA. As DNA damage accumulates, the cancer cells are overwhelmed with mutations and dysfunctional DNA leading to cell death.
The history of radiotherapy ran parallel to the efforts in understanding the molecular and genetic basis of cancer until combination therapy showed how combining medical physics with pharmacology could yield better results. Radiotherapy finds its origins in the 1890s when, shortly after the discovery of X-rays, radiation was used to treat some cancer patients successfully. In the first half of the 1900s, progress was made in defining radiation dosage and treatment regimen, and cobalt therapy machines were adopted between 1950 and the early 1980s. These machines emit gamma rays, which allow the treatment of deeper cancers. Subsequently, linear particle accelerators replaced cobalt units as they can produce higher energy beams without the need for a radioactive source inside them. Starting in the 1970s, radiotherapy saw the advent of more sophisticated techniques that included computerised tomography and integrated computerised image therapy that increased accuracy and efficacy thanks to three dimensional modelling of the tumour mass. Radiotherapy is often combined with chemotherapy, together with certain types of the afore-mentioned targeted therapies. In preclinical settings, several of these agents have been demonstrated to enhance the effect of radiotherapy making cells more susceptible to radiation damage. However, the safety profile of combination therapy is still unknown for many of these agents.
An alternative approach to combining radiotherapy and targeted therapy was offered by gold nanoparticle (AuNP)-based therapeutics. Gold nanoparticles boast several chemical and physical properties that make them suitable for cancer treatment. Among the physical properties of AuNPs, localised surface plasmon resonance, radioactivity and high X-ray absorption coefficient allow AuNPs to absorb incident photons and convert them to heat to destroy cancer cells (photothermal therapy). On the chemistry front, AuNPs can form stable chemical bonds with sulphur and nitrogen-containing groups, which allows AuNPs to bind to a variety of biological molecules and drugs. Taken together these properties allow AuNPs to be targeted to the cancer by coupling them to antibodies that recognise proteins, often oncogene products, on the surface of cancer cells. However, there are still limitations to their application in the clinic. For instance, phototherapy utilises light to excite the AuNPs, but it is strongly limited by the fact that light cannot reach more than a few centimeters under the skin surface, making some cancers unreachable. Using deeper-penetrating radiation such as gamma-rays could be a solution: gamma rays would excite core electrons near the atomic nucleus of gold and electrons may be released by a so-called Auger de-excitation process, which would then cause DNA damage and cell death. However, the optimal size of AuNPs to achieve this effect is still subject to debate. Further research in the field of medical physics holds promises in leveraging new technologies coupled with current knowledge of cancer biology, oncogene addiction and targeting to yield more efficacious treatments.
What’s next? To AI and beyond | Having explored and celebrated the discovery of oncogenes and the terrific knock-down effects that this discovery has had in the field of cancer treatments, it is now time to project into the future. What will oncogene research and related scientific efforts reveal? The initial oncogene findings were fundamental in revealing cancer as a genetic disease. They also appeared to explain cancer in simple terms with changes in one, or at most a few genes, to yield novel and specific therapeutic targets. However, careful cancer genome analyses, as part of the cancer genome project, have uncovered an unexpected multitude of genetic changes in all cancers, revealing complex mutational landscapes. Increasing complexity can also be found in our investigation of the role of oncogenic genes’ products. All these proteins show multiple activities, giving rise to diverse and specific cancer phenotypes. A complete molecular understanding of how these activities cause and maintain cancer remains a challenge, but striving to characterise cancers and go beyond an initial level of understanding oncogenes can lead to better diagnoses and treatments. Future revelations in the field of oncogene research may come from Big Data science and Artificial Intelligence (AI) which hold promise in addressing some of the challenges and questions that keep haunting oncologists and researchers.
Characterising cancers often remains a challenge, especially when fast sequencing of the whole cancer genome is not quickly available and extracting knowledge that can guide treatment is difficult. Oncogenes can be detected and combined with a more holistic range of cancer prognosticators, with AI bringing the data together for faster and more accurate treatment decisions. Recently, a convolutional neural network demonstrated efficacy in interpreting cancer biopsy slides by measuring not only the appearance of cancer cells but going beyond human vision to predict the driver mutations and wider transcriptomic changes. The algorithm was able to identify patterns and correlations between mutations and histological appearance, namely the aspect of the cancer tissue under the microscope. More than 17,000 slide images from 28 cancer types were analysed and correlated to the matched genomic, transcriptomic, and survival data of the patients who suffered from those cancers. While the researchers admitted that the number of associations between histopathology and molecular traits is remarkable, this AI approach is still lacking the same accuracy of genetic and transcriptomics testing. However, better algorithms hold promises for future diagnostics and for pointing out correlations between genetics, transcriptomics and cancer behaviour that can be used as a starting point for further experiments. Such experiments may reveal new mechanisms behind certain cancer features, in the same unexpected way a chicken virus ended up revealing one of the key principles of cancer biology. Ultimately, the fuller-picture AI can help establish may lead to faster and more accurate diagnosis and more personalised management options. Therapeutic development may then evolve to a multi-omic approach, targeting the full cancer signature rather than a single gene, so what started from oncogenes may become a system-wide treatment
Benedetta Spadaro and Harry Bickerstaffe are
Therapeutic Sciences MPhil students at St John's and Homerton colleges, respectively. Artwork by Marzia Munafo and Nataliia Kuksa.