The 2025 Pilot Award Awardees

Myelodysplastic syndrome (MDS) is a disorder in which the bone marrow stops producing blood cells properly. Patients often require transfusions and feel fatigued from inflammation. There is also an increased risk of developing leukemia, a blood cancer. Although there has been considerable progress made in identifying the gene mutations that cause MDS, how these mutations lead to impaired blood cell production and leukemia is still not clear. In addition, few new treatments have been developed over the past few decades. A major obstacle has been the lack of a good human model system for experiments in the lab. Use of animals, such as mice, don’t fully mirror the human disease and aren’t amenable to screening large numbers of drug candidates. Bone marrow cells taken directly from MDS patients are limited and don’t grow well in the petri dish. If they do grow, the petri dish fails to recapitulate the complex microenvironment within the bone marrow. Organoids are small structures grown from stem cells of different tissues. As they grow, they self-organize into the different cell types and mimic the architecture and microenvironment of that tissue. They have been used to study the intestine, lung, kidney, and brain for over a decade. They can be grown for months at a time and have been used successfully to identify new drugs. The first organoids to study the bone marrow were only reported in 2023, by one of our collaborators. These mimic the human bone marrow in many ways and are amenable to genetic experiments and drug screening. We hypothesize that BM organoids can be developed to study MDS. In this pilot study, we plan to generate human BM organoids that harbor two common MDS-associated gene mutations as proof-of-principle. We will analyze them to see how well they mirror the disease found in patients using a variety of techniques. Lastly, we will treat them with a drug that is known to affect one of the mutations as a test of whether the system can be used to test drugs. If successful, future work would apply this to more MDS-associated gene mutations, use the system to further understand how the mutations impair normal blood cell production and lead to leukemia, and to screen for new drugs to treat MDS.

The most common form of myelodysplastic syndromes (MDS), which affects 20% of patients with MDS, is characterized by the formation of tumoral cells called ringed sideroblasts. Because of these cells, the disease is named MDS with ringed sideroblasts (MDS-RS). The cause of this type of MDS is an alteration in the DNA that changes the function of a protein called SF3B1. MDS-RS has a low propensity for leukemic transformation, but patients with MDS-RS have severe anemia and require regular blood transfusions. Because of these continuous transfusions, patients with MDS-RS can die from iron overload in the liver and other adverse outcomes.
The only therapy approved for the treatment of MDS-RS patients who require transfusions and are ineligible for or have no response to drugs that stimulate red blood cell formation is luspatercept. However, luspatercept has a response rate of less than 40%, which underscores the need for alternative treatment options that can alleviate anemia in these patients. 
In our previous study, we found that the EIF2AK1 protein is responsible for anemia in these patients. Therefore, we hypothesize that pharmacologically blocking EIF2AK1 can overcome aberrant red blood cell formation and thus anemia in patients with SF3B1-mutant MDS-RS. To test this hypothesis, we developed two drugs targeting this protein. 
In the proposed work, we will functionally validate these candidates using cellular assays that we have developed over the last 4 years. The successful completion of this study will enable the development of the first inhibitor of EIF2AK1 activity. 
The proposed work has implications for the development of therapies to achieve long-lasting hematological responses in transfusion-dependent patients with MDS-RS. Given that EIF2AK1 inhibition is one of the most promising therapeutic approaches for sickle cell anemia to date, the results of our work may have a broad spectrum of clinical applications.

While the diagnosis of myelodysplastic syndrome (MDS) has been made easier due to the identification of recurrent genetic alterations, morphologic evaluation of the bone marrow (BM) is still necessary for an MDS diagnosis and remains challenging. In this project, we will determine if artificial intelligence/machine learning (AI/ML) approaches applied to digital images of BM biopsies (BMBx’s) can improve our ability to diagnose and prognosticate outcomes in MDS. We have collected the BMBx’s of newly diagnosed MDS patients and their clinical and morphologic mimics, as well as their associated genetic and clinical data. Using these digital images of the biopsies, we will train an AI/ML model and test its ability to accurately distinguish MDS from its morphologic mimics as well as predict clinically relevant features in MDS, including gene mutations, risk of disease progression, response to therapy, and overall survival. We have assembled large numbers of BMBx’s for our MDS and control groups, thereby ensuring the opportunity to build robust models, and these findings will be independently validated using a large group of patient samples from a second institution. If such AI/ML-based digital image evaluations of BMBx’s can successfully distinguish MDS from its clinical mimics and/or predict clinical outcomes, these models have the potential to fundamentally change how BMBx’s are evaluated in the clinical setting as well as promote health equity in under-resourced areas. Our investigative team’s collective expertise in hematopathology and AI/ML approaches positions us well to perform these studies.

Myelodysplastic syndromes (MDS) are blood cancers caused by mutations in bone marrow stem cells that impair blood cell production and may progress to leukemia. It is well known that mutations occurring with age and affecting bone marrow stem cells give rise to MDS. At the same time, studies in animal models suggest that the environment within the bone marrow surrounding MDS mutant cells also contributes to MDS development and may impact response to therapy. Despite these findings, little is known about how MDS mutant cells interact with their surrounding environment in the bone marrow of MDS patients.

We recently performed a comprehensive analysis of bone marrow biopsies from patients with MDS using a novel sequencing technology termed “spatial transcriptomics”. This technology enables sequencing of individual cells within intact tissue sections, allowing us for the first time to understand the identity and gene expression of each cell in the bone marrow while simultaneously determining its exact location. Importantly, we combined these data with single cell sequencing of aspirated bone marrow cells from MDS patients to map the location of mutant cells and clonal immune cell populations in patient bone marrow biopsy sections. 

Using this approach, we discovered that MDS patients harbor distinct clusters of immune cells surrounding MDS bone marrow stem cells. Based on these data we hypothesize that MDS mutant cells may trigger a local immune response within the bone marrow and may be reacting to MDS-specific antigens. This proposal addresses the above hypothesis through two aims. In Aim 1, we will use spatial transcriptomics to map the precise location of MDS mutant cells in the bone marrow, analyze their interactions with neighboring cells, and determine how their gene expression patterns differ across patients with specific mutations. In Aim 2, we will identify dominant B cell clones in the bone marrow of MDS patients—indicating a potential response to disease-specific targets. We will reconstruct the antibodies produced by these B cells and use mass spectrometry to identify the proteins they bind to in MDS patients.

This work will provide critical insight into how the immune system recognizes and responds to MDS mutant cells and enable characterization of potential antibodies to treat MDS patients. Ultimately, our goal is to define the spatial organization of disease-initiating cells in MDS and to identify novel targets of anti-tumor immunity in MDS.

MDS is due to genetic changes acquired by the diseased stem cells which normally are responsible for blood cell production the marrow. There are many types of genetic defects in MDS: tiny change in the genetic code called “mutations” determine the clinical course of the disease. SETBP1 is one of the genes that is frequently affected and when it happens the resultant disease progresses quickly and is difficult to treat. There are currently no specific treatments for this type of MDS. In this proposal we will tests potential drugs which we have already developed for their ability to inhibit this specific type or MDS but also other types in which SETBP1 may play a role because we have learned that it is present at high levels in a significant portion of patients with MDS event if the genes are not affected by mutations. Funding this project will directly advance the development of an entirely new drug class that may be helpful in treating patients with MDS.