Powered By Bing

Mullighan Lab: Research

The goal of research in the Mullighan laboratory is to advance cure rates for acute lymphoblastic leukemia and related cancers by identifying the genomic drivers of leukemia and treatment failure, and translate these findings into new diagnostic and therapeutic approaches by mechanistic interrogation and preclinical modeling. The lab provides an integrated research environment that encompasses genomic discovery from very large cohorts of primary human leukemia samples, experimental modeling using mammalian cell line, engineered mouse modeling, human CD34+ HSC modeling, and preclinical modeling using xenografts. Many of our discoveries have influenced the field of cancer genomics in general, and have been translated into new diagnostic techniques and clinical trials worldwide. The lab has made many of the seminal discoveries that have transformed our understanding of the genetic basis of leukemia and related cancers. Several areas of accomplishment and ongoing research are described below.

The laboratory provides an exceptional environment for training in experimental and computational biology relevant to acute leukemia, and prospective postdoctoral fellows with experience and/or interest in joining the laboratory are strongly encouraged to contact Dr. Mullighan. Funded training positions are available, and the majority of trainees have secured competitive training fellowships, publications in leading journals, and positions as scientists or independent faculty.

We make all reagents, models and data available to the research community. ALL genomic data may be mined here. Explore and request ALL Patient Derived Xenografts from our repository of over 200 samples (PROPEL). contact Aman Seth, preclinical project manager for the Hematological Malignancies Program, with any questions regarding PROPEL.

Genomic discovery: revising the molecular classification of ALL

tSNE plot

Fig 1: tSNE analysis of RNA-seq data of 1988 ALL transcriptomes (Gu et al, submitted)

The lab began in the field of genomic interrogation of ALL and continues to lead studies defining the genetic basis of ALL in children and adults, as well as a number of related high risk leukemias (see recent reviews on ALL, genetics of ALL, and Ph-like ALL). We continue to be engaged in several large scale discovery genomic projects that examine germline and somatic genetic changes from sub-whole genome (exome, RNA-seq, SNP array) and increasingly, WGS approaches. Using “first generation” genomic approaches, we identified many of the focal, submicroscopic deletions and mutated pathways characteristic of ALL (Mullighan et al., Nature 2007), and began to shed insight into the relationship of genetic lesions in disease lineage (e.g. deletion of IKZF1 and lymphoid lineage in BCR-ABL1 leukemia, Mullighan Nature 2008) and relapse in B-ALL (Mullighan et al., N Engl J Med 2009). These studies, others described below, and mouse modeling, showed the central importance of alteration of lymphoid transcription factor genes in the pathogenesis in B-ALL. With many of the studies examining germline alterations in ALL, and clonal evolution, we have described the sequential acquisition of genomic alterations in the pathogenesis of ALL.

We have described the genetic basis of important subtypes of ALL in detail, including hypodiploid ALLĀ  (Holmfeldt et al., Nat Genet 2013) and DUX4-rearranged ALL (Zhang et al., Nat Genet 2016), which elucidated a novel mechanism of sequential deregulation of transcription factor deregulation in ALL.

An important advance has been the understanding that while gene expression profiling remains an important unifying classifier of ALL, many new subtypes are characterized not by one single genetic alteration such as a chromosomal rearrangement but by diverse alterations converging on a single gene, or multiple genes (Gu et al., Nat Commun 2016 and submitted).

Ongoing studies are performing large scale genomic analyses incorporating whole genome, exome, and transcriptome sequencing of unbiased cohorts of over 2000 cases of childhood and adult ALL to fully understand the constellations of germline and somatic genomic alterations in leukemogenesis, their association with treatment outcome, and to fully resolve the genetic basis of enigmatic cases of ALL that have lacked known driver lesions. These studies have also formed the basis for ongoing studies modeling newly identified alterations in leukemogenesis.

Ph-like ALL

Ph-like ALL

Fig 2: Core kinase signaling pathways deregulated in Ph-like ALL (from Iacobucci and Mullighan, J Clin Oncol 2017;35:975)

Our laboratory co-discovered BCR-ABL1-like, or Ph-like ALL in 2009 (Mullighan et al NEJM 2009). This form of leukemia is one of the most common and high risk subtypes of ALL most common in older children and adults. Our laboratory has led the key efforts to describe the remarkably diverse genomic landscape of Ph-like ALL (Roberts et al Cancer Cell 2012 and NEJM 2014), describe the prevalence of Ph-like ALL across the age spectrum (Roberts et al J Clin Oncol), define the potential for inhibition of kinase signaling in ALL (Roberts et al Blood Adv 2017), and establish new mouse models of Ph-like ALL that show remarkable sensitivity to kinase inhibition (e.g. ETV6-NTRK3 ALL, Roberts et al Blood 2018 in press). Ph-like ALL is one of the most clinically important forms of ALL due to its high risk of treatment failure and potential for therapeutic targeting with FDA-approved tyrosine kinase inhibitors such as imatinib and dasatinib. These findings have had global impact, leading to revision of the molecular classification of ALL but the World Health Organization, multiple trials prospectively evaluating the potential of TKI therapy (e.g. Total Therapy XVII), and naming of the discovery of Ph-like ALL as one of the most important advances in the field of hematology in 2015.

Characterization of high risk ALL

The laboratory has a major interest in defining the genetic basis of poorly understood, high risk forms of leukemia and translating these findings to new mechanistic and therapeutic approaches.

Lineage ambiguous leukemias include those in which the leukemia cells exhibit myeloid and lymphoid features. One such leukemia is early T-cell precursor leukemia, so named as it has features of the earliest stage of human thymic T cell development. We defined this form of leukemia in 2009 (Coustan-Smith et al. Lancet Oncology) and our whole genome sequencing study that defined the genomics of ETP-ALL was the first genome sequencing study of any pediatric malignancy (Zhang et al. Nature 2012). This study defined the common mutations in genes regulating hematopoietic development, signaling and chromatin remodeling as hallmarks of this disease, and the cell of origin to be a hematopoietic progenitor. These results drove our interest to characterize the basis of other lineage ambiguous leukemias, and our recent large scale study of mixed phenotype acute leukemia showed that the lineage plasticity of this type of leukemia was due to the nature of the founding alterations (e.g. WT1 mutations in T/myeloid leukemia and ZNF384 rearrangements in B/Myeloid leukemia) and the stem cells in which they arise (Alexander et al, 2018 and in press). These studies have had important implications for leukemia diagnosis and classification, and have led to multiple ongoing experimental studies examining cell of origin and the roles of these alterations in leukemia pathogenesis.

Acute erythroleukemia. This form of leukemia is characterized by excessive erythroblasts in the bone marrow, a degree of similarity to AML and myelodysplasia, and dismal outcome. Our interest was stimulated by the observation of lineage-inappropriate rearrangements and “hijacking” of the erythropoietin receptor in Ph-like ALL (Iacobucci et al. Cancer Cell 2016). We have characterized the genomic features of a large cohort of childhood and adult AEL, shown that these are distinct from AML and MDS, and that there are 4 key groups defined by genomic changes, gene expression profiles, and varying outcome. We have shown the potential for therapeutic targeting of multiple pathways including NTRK signaling and epigenomic modifications, and using genome editing of mouse HSCs, have demonstrated that roles of specific combinations of genomic aberrations in defining disease lineage and phenotype (Iacobucci et al 2016, submitted and in preparation).

Clonal evolution and mechanisms of treatment failure in ALL

Clonal Evolution

Fig 3: schema of clonal evolution from diagnosis to relapse in ALL (from Mullighan et al Science 2009)

The lab has led many of the studies that have examined the relationship of genetic variegation, clonal evolution in relapse in ALL, and studies the mechanisms by which relapse-enriched mutations drive drug resistance. We published a seminal study showing that in most cases of ALL, there is partial sharing of the genetic alterations between the predominant diagnosis and relapse clones, and that relapse-enriched mutations could be backtracked to minor clones in the majority of cases (Mullighan Science et al 2008). This indicates that the relapse-fated clone is commonly present at diagnosis, and shares a common ancestral origin with the diagnosis predominant clone. These principles have now been confirmed in many, if not all cancer types. We have performed large scale integrated genome-wide sequencing, xenograft and drug treatment analyses large cohorts of relapsed ALL to fully elucidate the relationship of mutational status, transcriptional perturbation, clonal structure and drug resistance (Ma et al. Nat Commun 2015 and submitted). We reported CREBBP to be the most common target of sequence mutation at relapse (Mullighan et al, Nature 2011). Ongoing studies are characterizing the mechanism of drug resistance resulting from CREBBP mutations, performing functional genomic screens in mouse and human cells to identify relapse drivers, integration of single cell approaches to clonal structure analyses, and using genomic approaches to study the mechanisms of resistance to immunotherapy.

Germline predisposition to ALL

Germline predisposition to ALL

Fig 4: Role of IKZF1 as a cancer predisposition gene (Churchman et al Cancer Cell 2018)

The laboratory has identified many of the genes that predispose to the development ALL predisposition genes, and studies that have characterized the leukemogenic effects of these alterations. These include germline mutations in TP53 which are characteristic of low hypodiploid ALL (Holmdeldt et al. Nat Genet 2013). We reportedĀ  inherited mutations in the B lymphoid transcription factor PAX5 in ALL with dicentric/isochromosome 9p (Shah et al. Nat Gene 2013), and have generated kockin PAX5-mutant mouse models of ALL using CRISPR/Cas9 technology (Gu et al, submitted). We recently showed that IKZF1 germline mutations predispose to both familial and sporadic ALL, and that these inherited variants markedly influence IKZF1 function and drug resistance (Churchman et al. Cancer Cell 2018). Our ongoing studies are examining the genetic basis of familial ALL as part of a collaborative study funded by an NCI X01 Gabriella Miller grant), and dissecting the mechanistic basis of the characteristic IKZF1 mutations observed in familial ALL.

The leukemia microenvironment and drug resistance

Leukemia microenvironment

Figure 5: Schema for mechanism of action of IKZF1 alterations promoting drug resistance in ALL, and potential therapeutic targeting.

The lab is using several approaches to understand the role of interactions of leukemia cells with the microenvironment as a mechanism of drug resistance. We described the marked enrichment of alterations of IZKF1 (encoding the lymphoid transcriptional regulator Ikaros) in high risk B-ALL (Mullighan Nature 2008 and Mullighan et al., N Engl J Med 2009), and subsequently showed that these alterations by resulting in acquisition of a stem cell-like phenotype, and aberrant cell-stromal adhesion that results in bone marrow niche mislocalization (Churchman et al, Cancer Cell 2015 and review). This work identified rexinoids and FAK inhibitors (Churchman et al. JCI Insight 2016) as agents to overcome this deregulation and drug resistance. We are now extending these studies to examine the general importance of interaction of ALL cells with the niche as a mechanism of quiescence and resistance, using in vitro and in vivo imaging, multiomic and humanized niche models (unpublished data).

Collaborations, funding and resources

The laboratory is closely integrated with the St Jude Hematological Malignancies program and many of the discoveries from the laboratory have been incorporated into clinical trials at St Jude and world-wide. We have strong links with pediatric and adult cooperative groups including the Children’s Oncology Group, the NCT TARGET initiative, MD Anderson Cancer Center, The ACRIN Eastern Cooperative Oncology Group, and the Alliance Cancer and Leukemia Group B.

Research is supported by multiple sources, including an NCI Outstanding Investigator Award, Leukemia and Lymphoma Society SCOR and TRP grants, and foundation grants.