Answers are grounded in this paper and cite their sources.
Cancer as a genetic disease begins with genetic lesions in proto-oncogenes, key regulators of cellular physiology. Discoveries by Riccardo Dalla-Favera, MD, and his team have shaped the modern cancer genetics paradigm, from cloning the first human proto-oncogenes and their cancer-associated variants to elucidating mechanisms of genetic alterations, and mapping the biological processes by which the genetic lesions cause lymphoma. “Most human diseases are cured when you know the mechanism,” Dalla-Favera believes. Toward this aim, his seminal contributions have led to diagnostic and therapeutic advances in lymphoma. Dalla-Favera's discoveries have been recognized by this year's AACR Award for Outstanding Achievement in Blood Cancer Research. As the director of the Institute for Cancer Genetics and professor at Columbia University, New York, Dalla-Favera continues pursuing fundamental questions regarding the genetic mechanisms of cancer through the lens of lymphoma. He shares his perspective on the history, current state, and future avenues of the field in conversation with Blood Cancer Discovery's Tanya Bondar.What was known about causes of cancer when you started your research?It was clear that cancer was a genetic disease. For example, exposures that modify DNA chemically or physically were known to cause cancer in animals and humans. But we had no idea about which genes are responsible until the era of gene-cloning technology in the mid-70s. That allowed studying viral oncogenes—the genes of retroviruses that can cause cancer. The turning point for me was the paper showing that this viral gene was derived from host (chicken) genes—a discovery that gave the Nobel prize to Bishop and Varmus (1). My project was simple: if chickens have proto-oncogenes, humans must have them as well. And so, by using as probes transforming genes from a primate oncovirus, we cloned several human proto-oncogenes (2). Importantly, these were normal human genes, with exons and introns. The correct name is proto-oncogenes, because they're normal. The name “oncogene” is correct for their modified cancer-causing forms.Among them was c-MYC. Is it a coincidence that the proto-oncogene now known as a culprit across most human cancer types was cloned by a hematologist?As I was trained as a hematologist, I searched for proto-oncogenes involved in blood cancer. We used as a probe a viral oncogene of the avian myelocytomatosis virus. Its name, v-myc, suggested a role in myeloid cells. This is how we cloned human c-MYC (3).It is remarkable that c-MYC's name comes from “myeloid.” Engraved in the name of this powerful proto-oncogene is an example of how blood cancer research is leading the way.Riccardo Dalla-Favera's career goal has been “to dissect the genetics of cancer via B-cell lymphoma as a way to cure this disease.”I think hematology has pioneered research across fields in many ways. For example, the concept of stem cells: the first purified stem cell was hematopoietic. And hematology has always led the way on lineage determination and mechanisms of differentiation, partly because bone marrow, and especially blood, are easily accessible.What were the first insights about how a proto-oncogene turns into an actual oncogene?We took advantage of the well-known notions by Janet Rawley and other cytogeneticists that hematologic malignancies had extraordinary recurrence of chromosomal translocations that were the same in every cell of a given tumor, and specific translocations were associated with specific disease types. And since we had a human gene related to oncogenesis, the simple thinking was: let's map this gene onto the chromosomes and see if it might be involved in these specific translocations associated with human tumors. And luckily enough, our lab at NIH in collaboration with Carlo Croce, by a somatic cell hybrid, mapped the MYC gene to a region of chromosome 8 that turned out to be exactly the region that was broken in Burkitt lymphoma—a relatively rare B-cell tumor type (4).In 1983, the junction between c-MYC and the immunoglobulin heavy-chain locus (IgH) was cloned (5). This demonstrated that the recombination placed the expression of c-MYC (at that point already defined as a growth-promoting gene) under heterologous control of constitutively active regulatory domain of a gene within the IgH locus. From that, mouse models by Gerry Adams's group and our work in human cells demonstrated that deregulated expression of c-MYC mimicking a translocation would transform EBV-immortalized human B cells in vitro (6).Another insight came from the most common type of lymphoma, diffused large B-cell lymphoma (DLBCL). It started with an astute observation by Raju Chaganti, a cytogeneticist at Memorial Sloan-Kettering Cancer Center, that there is a recurrent breakpoint that recombines with several chromosomes, including the Ig locus. This is how BCL-6 came about (7). Its role in DLBCL became evident from its tissue expression: it is expressed specifically in the germinal center. This pointed to GC as the cell of origin of DLBCL.Several of these seminal papers were published in 1982–83, just months apart. What was the key unlocking this wave of discoveries?Gene sequencing and cloning technology enabled key discoveries in cancer biology. After MYC was cloned, thanks to the collaboration with Carlo Croce, it was mapped to the section of chromosome 8 that was involved in chromosomal translocations in Burkitt lymphoma. I then used Southern blotting and showed that the gene was rearranged and linked to IGH locus. Immunologists used human Burkitt lymphoma and mouse plasmocytoma cell lines as a way to study IGH rearrangements, and they were puzzled for a while that in some of these cell lines, the IGH locus was linked to an unknown piece of DNA. When sequencing technology came about, they found this piece was MYC. Thus, while I came from MYC to Ig, they came from Ig to MYC. c-MYC was exciting as the first example of a general principle. Two other major genetic findings the same year were mutant RAS (8, 9) and translocation of c-ABL proto-oncogene to Philadelphia chromosome (10), a poster child of precision therapy.From your works, we know about unique mechanisms of oncogenesis pertaining to B cells because of their lineage program, somatic hypermutation, and Ig rearrangement. How much of lymphomagenesis is due to these processes, and how much is shared with other cancers?Lymphomas have the same mechanisms as many other tissues, such as copy number variations and point mutations. And they also have particular biology of the tissue, the germinal center, where modification of DNA are physiological and unique for this cell type. So a number of genetic mistakes that generate lymphomas is due to particular biology of the cells from which they derive. However, the AID enzyme that generates somatic hypermutation and class switch recombination belongs to the APOBEC family of genes that are ubiquitously expressed and act as defenses against viruses. Findings in solid tumors suggest that abnormal activation of APOBEC may contribute to mutagenesis in cancer.Do you see a role of nongenetic factors in lymphomagenesis?I think in B-cell lymphoma the initiation event may be a genetic lesion of an epigenetic factor gene. The genetic change in a chromatin modifier cascades epigenetic changes. Epigenetics is very important in my view, especially as a therapeutic avenue, because it's pharmacologically targetable. Another downstream component is tumor microenvironment, which is least in part determined by the tumor. The other major factors include signals from the tumor microenvironment and evasion of antitumor immunity.The classic view of oncogenesis is a stepwise acquisition of genetic alterations in proto-oncogenes and tumor suppressors. If a cell has a higher rate of DNA damage—which B cells presumably do because of RAG and AID—are you surprised B-cell lymphomas are not more frequent than carcinomas?Indeed, B cells undergo programmed genetic changes that are useful immunologically but carry cancer risks. There is a sophisticated process of selection to minimize the risks. The germinal center is a transient structure tightly controlled in time and space. It's induced by antigen, lasts 10 days, and disappears. During that time, most of the cells in the germinal center die, and few surviving cells move out in different states—as plasma cell, memory cells.Could you estimate the minimum number of oncogenic steps from a normal B cell to lymphoma?That's a difficult question. If we look at human tumors, a typical DLBCL has around 100 mutations that are suspected to be of some biological significance, major or minor. We do not know how few are necessary.Mouse models with a number of genetic lesions affecting oncogenes or tumor suppressors often don't develop lymphomas. Can we say these lesions are not sufficient?We can say that this combination is not sufficient. But frequency of cooccurrence in human tumors doesn't necessarily mean cooperation of these genetic lesions in mouse oncogenesis. Altering c-Myc and p53 in any cell type results in tumors. Other gene combinations are oncogenic in cell type–specific ways—for example, the combination of chromatin modifier inactivation with Bcl-2 activation in lymphoma. And in other cases, two genes that are clearly involved in human cancer may not show any additive effect when combined in mouse models. The biology in naturally occurring tumor is very sophisticated in combining pathways.How have the definitions of “proto-oncogene” and “tumor suppressor” genes evolved to this day, and is it still a useful framework to describe genetic mechanisms of oncogenesis?30 years you would find a gene altered in cancer and say you found a new oncogene. Fortunately, it is not like this anymore because it's too superficial. I would at this point stay away from the word “oncogene” because it doesn't communicate that it's a normal gene that gets altered. Tumor suppressor is still a valid definition. Yet together they don't cover the whole spectrum. It's better to classify the effect on oncogenesis by function, such as proliferation, metabolism, extracellular environment.How would noncoding alterations fit in that classification?Characterization of oncogenic alterations in the regulatory domain of the genome is just emerging. On the one hand, there is enormous literature: large genome-wide association studies have identified noncoding regions as predisposing to certain tumors. The noncoding alterations occur not only somatically but also in the germline alterations. And I think the part in the germline is particularly interesting because they are linked to incidence of many tumor types. But mechanistic understanding has been lagging, because functionalization of the noncoding genome is difficult. Only now technologies to explore this are becoming available.Could prevention or earlier intervention be game changers in the lymphoma field?I do not see a screening biomarker of lymphomas, in the same was as prostate-specific antigen (PSA) for prostate cancer or colonoscopy for colon cancer. What do you screen for in lymphoma? Healthy individuals have plenty of clonal B cells. I personally think it will be difficult to prevent lymphomas because they are mistakes of physiology. There are precursor diseases, such as follicular lymphoma for DLBCL, monoclonal B-cell lymphocytosis for myeloma, and monoclonal B-cell lymphocytosis for chronic lymphoblastic leukemia. But to intervene at those stages we need better therapies, because it is hard to justify a toxic treatment when the risk of progression is low.How will lymphomas be treated 10–15 years from now?Without chemotherapy. The bar for a new therapy to compete with chemotherapy is currently high because, for example, 40% of DLBCL cases are cured by chemotherapy. So in clinical trials, new drugs are introduced as additions to chemotherapy. But in the future, instead of dropping a bomblike poison on the organism, we will have found a way to rationally attack the cancer such as by immunotherapy and targeted therapy.Could you give advice to young scientists on how to identify a project that's worth investing your time?My first advice is to focus. Rather than being driven by opportunity, choose a goal that defines your next 5 to 10 years and just pursue that. And ask yourself: if everything works, every dream in the experiment comes true, will I achieve something transforming? That's my philosophy. Make it to the textbook. My postdocs, my grandchildren will open a textbook and there will always be Myc.What was the goal you set for yourself as a postdoc?To dissect the genetics of cancer via B-cell lymphoma as a way to cure this disease. Most human diseases are cured when you know the mechanism. When I finish my career, if the genetics of B-cell lymphoma is much clearer and hopefully has contributed something to diagnosis and therapy, I'd be happy.
No takes yet. Share an insight, caveat, or question.
A 2023 study studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: