This transcript is software driven, please understand there may be errors. Should any inaccuracies or omissions be found, please notify transcripts@MedEdOTG.com for correction.
Hi. This is Dr. Rick Van Etten from the Chao Family Comprehensive Cancer Center at the University of California, Irvine. I'm going to talk about the clinical importance of mouse models of myeloproliferative neoplasms: What do they teach us, with a focus on what the clinical utility of these models has been.
Again, let's talk first about chronic myeloid leukemia or CML. These models have been very useful in several aspects in helping us to understand CML pathophysiology, with implications for treatment of patients. We all know that CML patients have different responses to tyrosine kinase inhibitor therapy, and the reasons are largely unknown, but this mouse modeling study showed that one of the things that influences the initial response to TKIs is the status of a protein called BIM. Now BIM is a pro-apoptotic protein, which triggers mitochondrial apoptosis. On the left panel, you can see that mice that either are missing one or two copies of the BIM gene have a greatly blunted response to the tyrosine kinase inhibitor imatinib, in terms of the ability to induce cell death in CML cells.
In the subsequent study by the group from Singapore on the table on the right, you can see that there is a polymorphism in the BIM gene that is particularly prevalent in East Asian cohorts. There is a very good correlation of initial response to imatinib, whether or not you carry this BIM deletion polymorphism in a population.
Another great area of interest in chronic myeloid leukemia are strategies to eliminate what we call the leukemic stem cells. The B cells are postulated to be the cause of persistence of disease and a relapse of disease when patients have their tyrosine kinase inhibitor therapy discontinued. Mouse models have been very important in identifying pathways that regulate leukemia stem cell survival and proliferation. Among these are the Wnt/beta-catenin pathway, the Hedgehog/Smo pathway, a pathway involving Alox5, which is a leukotriene metabolizing enzyme, and the PML gene, promyelocytic leukemia gene, which is involved in nuclear bodies and regulation of stability.
As the consequence of these discoveries, there are numerous clinical trials of targeted therapies, which are directed at these pathways. The goal of these trials is to try to eliminate minimal residual disease in CML, leading to permanent cure of the disease.
Now, these models have also been applied to the PH-negative model of myeloproliferative neoplasms, and one of the most valuable things in the next slide has been helping us classify the genetic mutations that we see in these MPNs and place them in different categories. On the left side, I have shown what we would call "Phenotypic driver mutations" in these diseases. These include mutations in JAK2, MPL, and, more recently, calreticulin and LNK/CBL. What I mean by a driver mutation is this is a mutation that is introduced into a hematopoietic stem cell in a mouse. It can recapitulate part, or all, of the myeloproliferative phenotype that we associate with the disease.
By contrast on the right, there's a group of genes that I call modifiers. These include TET2, IDH1 and 2, DNA methyltransferase 3a, and others. These genes, when they're reintroduced into mouse stem cells, do not recapitulate by and large the MPN phenotype, but instead may have other phenotypes. For instance, type 2 IDH1 and 2, and DNMT3-a, when introduced into mouse stem cells, all increased stem cells self-renewal, but did not cause myeloproliferative phenotype. Hence, they will be placed in a category known as modifiers.
Another thing where mouse models have been found to be useful is to try to answer the question of genotype, phenotype correlation. By that I mean the JAK2 V617F mutation is found in virtually every patient with polycythemia vera, but can also be found in about half of patients with essential thrombocythemia and myelofibrosis. Why might that be? There is persuasive evidence from mouse models that it is the mutant allele burden of JAK2 and its expression level that may influence the disease phenotype.
This is summarized on a figure from a review from Tony Green's lab, which shows that in the various mouse models that express JAK2 V617F, at different levels you find either an ET-like phenotype with predominantly elevated platelets, or a PV-like phenotype with predominant elevation of the hematocrit. I think mouse models have gone a long way toward explaining this clinical conundrum.
The other thing the mouse models have been good for, as I've mentioned, is showing the difference between the driver and non-driver mutations. As I've mentioned, mutations in TET2 and DNMT3a have been actually recently identified by clonal analysis as preceding the driver mutations when they're required in the human disease. Interestingly, the same mutations are often found as acquired mutations in elderly individuals who have clonal hematopoiesis. This fits perfectly with their ability to increase self-renewal of stem cells in the mouse models.
To summarize what we have learned about the epigenetic regulators in MPN pathogenesis is that mutations in these genes that encode epigenetic regulators are frequently found as modifiers. Many of these mutations as single lesions have very subtle effects on hematopoietic stem cells self-renewal. The molecular mechanisms are as yet unknown. But importantly, although there's great interest in trying to target these lesions, most of these mutations result in a decrease or loss of function, and hence, they may not be good targets for therapy.
To summarize, a current model for the pathogenesis of PH-negative MPNs through mouse models is that there may be a heterozygous mutation in JAK2, in calreticulin, or MPL as the initiating mutation that may all have an ET-like phenotype. With an increase in the JAK2 gene dosage, you may progress to polycythemia vera, and then after acquisition of additional genetic or epigenetic processes, move on to myelofibrosis.
Clinically relevant questions remaining on the MPN pathogenesis: One is of what etiology is so-called triple-negative MPN, lacking mutations in JAK2, MPL, or calreticulin? There's a lot of interest in what the pathobiology of calreticulin mutant MPN is. What treatment modalities can impact the natural history in the malignant clone in the MPNs? Agents under investigation in mice and in humans include JAK2 inhibitors and interferon alpha.
Then, lastly, can we therapeutically exploit the mutations and epigenetic regulators in MPNs?
This completes the discussion of the clinical importance of mouse models of MPN, and I'd like to thank you all for your attention.