Over the last 10+ years, genome re-sequencing in AML has led to the identification of novel recurrent mutations in genes previously not implicated in AML pathogenesis including: IDH1/2, DNMT3A, TET2, and others. Importantly, these studies demonstrated that most cases of AML are associated with mutations in multiple genes, often occurring with different allele frequencies, suggesting a complex clonal architecture and developmental history. These findings raise many important questions, particularly how AML develops from normal hematopoietic cells and how multiple mutations accumulate in a single clonal lineage of hematopoietic cells. We proposed a model in which serial acquisition of mutations occurs in self-renewing HSCs, and investigated this model through the analysis of de novo AML and patient-matched residual HSCs. Using exome sequencing, we defined mutations present in nearly 40 individual AML genomes, and screened for these mutations in the residual HSCs. We identified several mutations present in these cells, identifying them as pre-leukemic mutations. Finally, through single cell analysis, we determined that as we hypothesized, a clonal progression of multiple mutations occurs in HSCs (Figure 2). We have extended these studies to serial patient samples demonstrating that pre-leukemic HSCs persist in clinical remission, and that there are multiple clonal paths to relapse, some of which may involve further clonal evolution of pre-leukemic HSCs. Ongoing work is focused on investigating: (i) whether the order of pre-leukemic and leukemic mutations is critical for the leukemic phenotype; (ii) the effect of pre-leukemic mutations on HSC function and chemosensitivity; and (iii) the contributions of pre-leukemic HSCs to AML relapse.

References

1. Corces MR, Buenrostro J, Wu B, Greenside PG, Chan SM, Koenig JL, Snyder MP, Pritchard JK, Kundaje A, Greenleaf WJ, Majeti R*, and Chang H*. Lineage-specific and single cell chromatin accessibility charts human hematopoiesis and leukemia evolution. Nature Genetics, 48: 1193-1203 (2016)
2.  Corces-Zimmerman MR, Hong W, Weissman IL, Medeiros BC, and Majeti R. Preleukemic mutations in human acute myeloid leukemia affect epigenetic regulators and persist in remission. P.N.A.S., 111: 2548-2553 (2014)
3.  Jan M, Snyder TM, Corces-Zimmerman MR, Vyas P, Weissman IL*, Quake SR*, and Majeti R*. Clonal evolution of preleukemic hematopoietic stem cells precedes human acute myeloid leukemia. Science Translational Medicine, 4: 149ra118 (2012)Clonal evolution of preleukemic hematopoietic stem cells precedes human acute myeloid leukemia. Science Translational Medicine, 4: 149ra118 (2012)

Precise and efficient manipulation of genes is crucial for understanding the molecular mechanisms that govern human hematopoiesis and for developing novel therapies for diseases of the blood and immune system. Current methods do not enable precise engineering of complex genotypes that can be easily tracked in a mixed population of cells. We described a method to multiplex homologous recombination (HR) in human hematopoietic stem and progenitor cells and AML cells by combining rAAV6 donor delivery and the CRISPR/Cas9 system delivered as ribonucleoproteins (RNPs). In addition, the use of reporter genes allowed FACS-purification and tracking of cells that had multiple alleles or loci modified by HR (Figure 3). We believe this method will enable broad applications not only to the study of human hematopoietic gene function and networks, but also to perform sophisticated synthetic biology to develop innovative engineered stem cell-based therapeutics.

We are now using these methods to investigate the effects of various mutations identified in pre-leukemic HSCs and myeloid malignancies on normal human hematopoietic stem and progenitor cells (HSPCs). Ongoing projects involve the investigation of mutations, including disrupted alleles and canonical point mutations, in DNMT3A, TET2, ASXL1 (Figure 4), RUNX1, IDH1/2, SRSF2, JAK2, FLT3, and NPM1. Next step questions involve investigating the mutation effects in young versus old human HSPCs and the cell autonomous and cell non-autonomous effects of combinations of multiple mutations.

References

1. Bak RO*, Dever DP*, Reinisch A*, Cruz Hernandez D, Majeti R*, and Porteus MH*. Multiplexed genetic engineering of human hematopoietic stem and progenitor cells using CRISPR/Cas9 and AAV6. eLife, 6; pii:e27873 (2017)Leukemia-associated cohesin mutants dominantly enforce stem cell programs and impair human hematopoietic progenitor differentiation. Cell Stem Cell, 17: 675-688 (2015)

Several methods have been developed to profile the epigenomes of rare cellular populations, enabling the identification of regulatory elements in progenitor populations in human hematopoiesis. Using the Assay for Transposase-Accessible Chromatin and sequencing (ATAC-seq), we defined the chromatin accessibility and transcriptional landscapes in 13 human primary blood cell types that span the hematopoietic hierarchy (Figure 5). Exploiting the finding that the enhancer landscape better reflects cell identity than mRNA levels, we enabled ‘enhancer cytometry’ for enumeration of pure cell types from complex populations. We identified regulators governing hematopoietic differentiation and further showed the lineage ontogeny of genetic elements linked to diverse human diseases. In AML, chromatin accessibility uncovered unique regulatory evolution in cancer cells with a progressively increasing mutation burden. Single AML cells exhibited distinctive mixed regulome profiles corresponding to disparate developmental stages. A method to account for this regulatory heterogeneity identified cancer-specific deviations and implicated HOX factors as key regulators of pre-leukemic hematopoietic stem cell characteristics. Thus, regulome dynamics can provide diverse insights into hematopoietic development and disease. Ongoing studies are investigating the role of epigenetic dysregulation across the disease spectrum of AML from diagnosis into relapse post-chemotherapy and exploring the hypothesis that disease relapse is associated with epigenetic clonal evolution.

References

1. Corces MR, Buenrostro J, Wu B, Greenside PG, Chan SM, Koenig JL, Snyder MP, Pritchard JK, Kundaje A, Greenleaf WJ, Majeti R*, and Chang H*. Lineage-specific and single cell chromatin accessibility charts human hematopoiesis and leukemia evolution. Nature Genetics, 48: 1193-1203 (2016)Isocitrate dehydrogenase 1 and 2 mutations induce BCL-2 dependence in acute myeloid leukemia. Nature Medicine, 21: 178-184 (2015)

Genomic studies of AML have defined the mutational spectrum of the disease and associations of specific mutations with each other and clinical outcomes. Such studies from large cooperative groups have defined the genomic landscape of AML and started providing guidance for prognostics and therapeutic options for some molecularly targeted therapies such as FLT3 kinase inhibitors and IDH mutant-specific inhibitors. We have expanded from these studies to examine the impact of clonal architecture, clonal and subclonal mutations and variants, on clinical outcomes and drug sensitivity. Our studies have determined that clonal architecture, inferred from analysis of variant allele frequency, provides additional information beyond just simple presence and absence of a mutation. In some cases, even the inferred order of mutations dictate distinct clinical outcomes.

We used direct single cell genotyping to investigate the ability of this newer technology to identify clonal variants and persistence across the treatment course from diagnosis to remission to relapse. We were able to detect clones responsible for relapse at the time of remission in several patients (Figure 6). In others, we were able to detect clonal hematopoiesis clones at the time of remission. It remains to be seen if this approach can be translated into a minimal residual disease assay. Ongoing studies are looking to use single cell genotyping data to refine the analysis of clonal structure and relationship to clinical outcomes at diagnosis. Additional studies are using single cell multi-omics including surface proteomics, gene expression, genotyping, and/or chromatin accessibility to refine the analysis of AML, stem cells, and minimal residual disease.

In separate studies, we have developed methods for barcoding CRISPR/Cas9 engineered cells. The TRACE-Seq method we developed incorporates a barcode into the HDR template that allows discrimination and tracking of single cell clones within a population of cells. This approach has applications for human gene therapy, as well as potential for clonal tracking in primary human cells in vitro and in vivo.

References:

1. Benard BA, Leak LB, Azizi A, Thomas D, Gentles AJ, and Majeti R. Clonal architecture predicts clinical outcomes and drug sensitivity in acute myeloid leukemia. Nature Communications, 12: 7244 (2021)
2. Sharma R*, Dever DP*, Lee CM*, Azizi A*, Pan Y, Camarena J, Koehnke T, Bao G*, Porteus MH*, and Majeti R*. The TRACE-Seq method tracks recombination alleles and identified clonal reconstitution dynamics of gene targeted human hematopoietic stem cells. Nature Communications, 12: 472 (2021)
3. Ediriwickrema A, Aleshin A, Reiter JG, Corces MR, Kohnke T, Stafford M, Liedtke M, Medeiros BC, and Majeti R. Single-cell mutational profiling enhances the clinical evaluation of AML MRD. Blood Advances, 4: 943-952 (2020)

Understanding the relative contributions of genetic and epigenetic abnormalities to AML should assist integrated design of targeted therapies. We generated induced pluripotent stem cells (iPSCs) from AML patient samples harboring MLL rearrangements and found that they retained leukemic mutations but reset leukemic DNA methylation/gene expression patterns. AML-iPSCs lacked leukemic potential, but when differentiated into hematopoietic cells, they reacquired the ability to give rise to leukemia in vivo and re-established leukemic DNA methylation/gene expression patterns, including an aberrant MLL signature (Figure 7). Epigenetic reprogramming was therefore not sufficient to eliminate leukemic behavior. This approach also allowed us to study the properties of distinct AML subclones, including differential drug susceptibilities of KRAS mutant and wild-type cells, and predict relapse based on increased cytarabine resistance of a KRAS wild-type subclone. This initial study illustrated the value of AML-iPSCs for investigating the mechanistic basis and clonal properties of human AML. Ongoing studies involve deriving AML-iPSCs from additional patient samples with more common mutations apart from MLL rearrangements and using CRISPR/Cas9 genome engineering in AML-iPSCs to explore the requirement for individual mutations in leukemia initiation and maintenance.

References:

1. Chao MP, Gentles AJ, Chatterjee S, Lan F, Reinisch A, Corces MR, Xavy S, Shen J, Haag D, Chanda S, Sinha R, Morganti RM, Nishimura T, Ameen M, Wu H, Wernig M, Wu JC, and Majeti R. Human AML-iPSCs reacquire leukemic properties after differentiation and model clonal variation of disease. Cell Stem Cell, 20:329-344 (2016)

Two genes are synthetically lethal (SL) when defects in both are lethal to a cell, but a defect in one is non-lethal. SL partners of cancer mutations are of great interest as pharmacological targets. We have developed both screening and computational approaches to identify pathways and genes that can be targeted in synthetic lethal interactions with recurrently mutated AML driver genes while sparing normal HSPCs. In our screening approach (Figure 8), we engineered AML cell lines to inducibly express mutant genes and infected them with an shRNA library expressed from barcoded lentiviruses. Deep sequencing of cells immediately after viral infection and several weeks later was used to identify potential synthetic lethal interactions. Thus far, we identified and validated several synthetic lethal interactions with IDH1/2 and are in the process of developing reagents for similar screens with additional AML mutant genes. In the case of IDH1/2, we found that inhibition or knockdown of BCL-2 is synthetic lethal with these mutations, and that a pharmacologic BCL-2 inhibitor exhibited potent and selective activity against IDH-mutant AML. In our computational approach, we developed MiSL (Mining Synthetic Lethals), a new algorithm that mines pan-cancer human primary tumor data to identify mutation-specific SL partners for specific cancers. We applied MiSL to 12 different cancers and predicted 145,891 SL partners for 3,120 recurrent mutations, including known mutation-specific SL partners. Comparisons with large-scale functional screens showed that MiSL predictions were enriched for mutation-specific SLs in multiple cancer types. We extensively validated a new SL interaction identified by MiSL between the IDH1 mutation and ACACA in leukemia using in vitro assays and patient-derived xenografts. Furthermore, we applied MiSL to pinpoint genetic biomarkers for tumor drug sensitivity.

References:

1. Sinha S*, Thomas D*, Chan S, Gao Y, Brunen D, Torabi D, Reinisch A, Hernandez D, Chan A, Rankin EB, Bernards R, Majeti R*, and Dill DL*. Systematic discovery of mutation-specific synthetic lethals by mining pan-cancer human primary tumor data. Nature Communications, 8: 15580 (2017)
2. Chan SM, Thomas D, Corces-Zimmerman MR, Xavy S, Rastogi S, Hong WJ, Zhao F, Medeiros BC, Tyvoll DA, and Majeti R. Isocitrate dehydrogenase 1 and 2 mutations induce BCL-2 dependence in acute myeloid leukemia. Nature Medicine, 21: 178-184 (2015)

Precursor B-acute lymphoblastic leukemia (B-ALL) is an aggressive malignancy of lymphoid precursors more common in children than adults. Several lines of evidence indicate that a block in normal B cell differentiation is key to the pathogenesis of B-ALL. During hematopoietic differentiation, stem cells give rise to lineage-restricted progenitors starting with the common myeloid (CMP) and lymphoid (CLP) progenitors. While they are lineage-committed during normal hematopoiesis, CLP maintain a latent myeloid potential that can be unmasked by ectopic cytokine receptor signaling. Investigation of the mechanism of this reprogramming implicated the transcription factor CEBP-alpha (CEBPA), as introduction of CEBPA initiated myeloid differentiation from CLP and pro-B cells. Thus, early lymphoid-committed progenitors can be reprogrammed to the myeloid lineage. Given that most human B-ALL cells exhibit a block in differentiation, we hypothesized that B-ALL cells can also be reprogrammed to the myeloid lineage. We determined that in vitro culture of primary B-ALL cells from many, but not all patients, with the myeloid-promoting cytokines IL-3, GM-CSF, and M-CSF produced cells expressing the myeloid markers CD14 and CD11b with the morphology and features of macrophages (Figure 9). Furthermore, these cells harbored clonal genetic abnormalities and clonal VDJ recombination events found in the original B-ALL, indicating that the reprogrammed myeloid cells shared a clonal original with the leukemia, and suggesting that culture with myeloid-promoting cytokines drove B-ALL cells to differentiate into myeloid-lineage cells resembling macrophages. Bioinformatic analysis revealed that B-ALL samples capable of myeloid reprogramming harbored a latent myeloid gene expression program, and introduction of the myeloid transcription factor CEPBA into such samples caused myeloid reprogramming in the absence of the cytokines. Finally, we found that these myeloid-reprogrammed ALL cells failed to engraft in vivo and cause leukemia in immunodeficient mice. As mature myeloid lineage cells have a limited lifespan, such reprogramming and trans-differentiation may represent an entirely novel therapeutic strategy for the treatment of B-ALL. Ongoing work is focused on further delineating the mechanism of cellular reprogramming and assessing the functions of these reprogrammed myeloid cells as candidate antigen presenting cells.

References:

1. McClellan JS*, Dove CG*, Gentles AJ, Ryan CE, and Majeti R. Reprogramming of primary human Philadelphia chromosome-positive B cell acute lymphoblastic leukemia cells into nonleukemic macrophages. P.N.A.S., 112: 4074-4079 (2015)

Conventional xenotransplant models have been used to study human normal and malignant hematopoiesis, but do not necessarily recapitulate human physiology since environmental factors provided by bone marrow (BM) niche cells may differ between mice and humans. We developed a novel mouse model bearing a subcutaneously accessible human bone ossicle formed by in situ differentiation of BM-derived mesenchymal stromal cells (MSCs) (Figure 10). These human ossicles contain a functional humanized BM niche that facilitates robust engraftment of normal human HSPCs and exhibits superior engraftment and expansion of primary AML. Interestingly, this model allows robust engraftment of PML-RARA positive acute promyelocytic leukemia (APL) and JAK2 V617F+ primary myelofibrosis (PMF), both of which fail to engraft in conventional models. Ongoing work includes manipulation of the bone marrow niche through knockdown methods to assess the impact on normal and leukemic engraftment.

References:

1. Reinisch A, Thomas D, Corces MR, Zhang X, Gratzinger D, Hong WJ, Schallmoser K, Strunk D, and Majeti R. A humanized bone marrow ossicle xenotransplantation model enables improved engraftment of healthy and leukemic human hematopoietic cells. Nature Medicine, 22: 812-821 (2016)