Mark Kamps - US grants

The objective of this proposal is to understand how expression of an altered form of PBX1, the prototype of a new homebox gene family, alters hematopoiesis and induces leukemia in man. I have shown that expression of a fusion transcript between the transcription factor gene, E2A, and the homebox gene, PBX1, is the genetic consequence of the t(1;19) translocation of childhood pre-B cell acute lymphocytic leukemia, and that the presumptive chimeric transcription factor encoded by this transcript produces myeloid leukemia in mice. Discovery of this new human oncogene prompts three distinct questions: first, what biochemical activities are required to make E2A-Pbx1 a transforming gene (Aims 1 and 2); second, what specific cellular events does E2A-Pbx1 alter to produce leukemia (Aim 3); and third, do other PBX-related genes control elements of hematopoetic differentiation (Aim 4). Aim 1: Determine the structural domains of E2A-Pbx1 required for transformation. Whereas E2A-pbx1 cDNA's are in hand, a full-length pbx1 cDNA must be isolated. The leukemic potential of normal Pbx1, an aminoterminal truncation of Pbx1 at the E2A- Pbx1 junction, and Pbx1 fusion proteins with other transactivation domains will then be examined in mice. Analysis of a shorter spliced variant of E2A-Pbx1 that lacks 7 Kda's of Pbx1 will permit further resolution of the functional requirements of Pbx1. Aim 2: Determine the transcriptional activity of Pbx1 and E2A-Pbx1. Recombinant Pbx1 proteins will be expressed in bacteria, purified, and then used to select a Pbx1 DNA-binding sequence from degenerate oligonucleotides or genomic DNA fragments. Transcriptional activities of Pbx1 and E2A-Pbx1 will then be assessed by cotransfecting a reporter construct (CAT), transcriptionally regulated by the Pbx1-responsive DNA sequence, with an expression plasmid, producing forms of Pbx1 or E2A-Pbx1. Transcriptional properties will then be correlated with transforming ability to determine a biochemical basis for transformation. Aim 3: Examine the cellular effects of E2A-Pbx1. The ability of E2A-Pbx1 to alter the growth and/or differentiation of granulocyte and granulocyte-macrophage, colony- forming cells from mouse marrow will be examined in vitro. The ability of E2A-Pbx1 to inhibit inducible differentiation of cell lines in vitro will also be tested. These results will reveal whether E2A-Pbx1 alone can mediate, the factor-dependent, differentiation-inhibited, in vivo phenotype of E2A-Pbx1-induced leukemias. Aim 4: Initiate studies to determine whether other PBX1 homologues are expressed during hematopoesis. The possibility that other Pbx1 homologs are expressed during hematopoetic differentiation and bind the Pbx recognition sequences will be addressed by cloning PBX homologues, using the known sequence identity between PBX1 and PBX2 to develop identification strategies. New cloned PBX CDNA'S will provide the basis to initiate studies on the role of PBX gene expression during hematopoeisis, as well as provide a basis for determining whether other PBX genes are involved in other types of human malignancy.

The purpose of this core unit is to provide the different project with oligonucleotides, synthetic peptides and purified recombinant proteins expressed in E. coli or in insect cells using the baculovirus expression system. The oligonucleotides will be used as linkers, PCR primers, probes for DNA binding assays, antisense inhibitors of gene expression and affinity ligands for the purification of DNA binding proteins. The peptides will be used as immunogens for preparation of anti-peptide antibodies, substrates for phosphorylation reactions, pseudosubstrate inhibitors of specific protein kinases and possibly phosphatases, ligands for protein purification and studying protein-protein interactions. Various transcription factors, such as cJun, T3Ralpha and E2A-Pbx1 and dominant acting inhibitors of them will be produced either in E. coli or in insect cells, purified and supplied to the various projects for microinjection experiments to study phosphorylation/dephosphorylation and interaction with other proteins. We will also produce various oncoproteins and kinases, such as Ha-Ras, vAb1, Erk1 and Erk2 to be used in microinjection experiments. This core will relieve the individual units from the routine chores of oligonucleotide, peptide and protein purification.

This proposal examines how the altered transcription of cellular genes by the E2A-Pbx1 oncoprotein blocks myeloid differentiation and causes fibroblast transformation. It tests whether E2A-Pbx1 induces transcription of known genes that regulate normal growth or of a novel population of genes that regulate differentiation and growth by an alternate mechanism. It also uses E2A-Pbx1-responsive cellular promoters as model systems to investigate transcriptional regulation by E2A-Pbx1 and by the normal cellular Pbx1 protein. E2A-Pbx1 is expressed as a consequence of the t(1;19) chromosomal translocation in pediatric pre-B ALL. It is comprised of the transactivation domain of the transcription factor E2A, and the DNA-binding domain of the new homeobox protein, Pbx1. E2A-Pbx1 is a nuclear phosphoprotein, and exhibits unique oncogenic properties. In a mouse bone marrow transplantation model, it induces acute myeloid leukemia, and infection of mouse marrow with E2A-Pbx1 virus in vitro results in the rapid outgrowth of GM-CSF-dependent myeloblasts. Therefore, E2A-Pbx1 blocks myeloid differentiation without altering growth-factor requirements. E2A-Pbx1 also transforms NIH3T3 fibroblasts, indicating that it may activate transcription of primary response genes. Specific aims 1-3 focus on 1) developing conditional mutants of E2A-Pbx1, 2) using them to clone E2A-Pbx1-regulated genes, 3) determining whether E2A-Pbx1-regulated genes directly control cell division or differentiation, and 4) using promoters of E2A-Pbx1-inducible genes to investigate the biochemical mechanism of transcriptional activation by E2A-Pbx1 as well as indirect transcriptional repression. The fourth specific aim proposes to identify mutations that strongly enhance fibroblast transformation by E2A-Pbx1 and to determine the biochemical property of E2A-Pbx1 that is affected. This will help establish a biochemical basis for transformation by E2A-Pbx1. The last specific aim proposes to use conditionally-transformed myeloblast cell lines to construct a model system to identify human oncogenes in AML that block myeloid differentiation. Because our major focus is the completion of aims 1-4, aim 5 will be developed as time permits. Accomplishing these aims will reveal how the pediatric leukemia protein, E2A-Pbx1, interferes with both growth and differentiation and the mechanisms through which oncogenes interfere with this regulation.

The objective of this proposal is to determine how the transcript for vascular endothelial growth factor (VEGF) is transiently stabilized by hypoxia in normal cells and becomes constitutively stabilized in a subset of tumor cells that highly overexpress VEGF. VEGF is a potent angiogenic peptide produced by solid tumors by at least two different mechanisms. First, concentrated zones of VEGF expression appear in hypoxic regions of solid tumors and are implicated in mediating hypoxia-induced angiogenesis. and growth. We find that upregulation of VEGF by hypoxia, TPA, and cyclohexamide is accompanied by a 3- to 6-fold stabilization of the VEGF transcript (from t 1/2 of 26-32 minutes to t 1/2 of 103-158 min), and that VEGF transcripts in human tumor cell lines that overexpress VEGF are constitutively stabilized 3- to 6-fold (1\2 of 113 to 181 min.) and exhibit little or no additional stabilization by hypoxia, TPA, or cycloheximide. This suggests that a normal signal transduction pathway involving an oxygen sensor regulates the stability of VEGF mRNA, and that mutations in this regulatory system cause persistent stabilization in some tumors. Our specific aims are designed to identify the primary sequences in VEGF mRNA that mediate basal instability and hypoxia-inducible stabilization, and to determine the molecular basis for constitutive stabilization of VEGF transcripts in breast carcinomas and other VEGF overexpressing tumors. Aim 1: Clone full-length, normal, 4.2 kb human and chicken VEGF transcripts from hypoxia-induced HBL 100 breast epithelial cells and chicken embryo fibroblasts. Evolutionarily conserved sequences represent potential elements that regulate message stability, and provide a framework for initiating mutagenic studies (Aim 3). Aim 2: Determine whether tumor cell lines exhibiting constitutive VEGF mRNA stabilization have mutations in their message stabilization pathways or in their VEGF transcripts. A slightly altered version of the normal human VEGF transcript, which can be distinguished from the endogenous transcript using RNase protection, will be expressed in tumor cell lines exhibiting constitutive VEGF message stabilization Basal and hypoxia- or TPA-induced stability will be measured for both the endogenous and exogenous VEGF mRNA's, using RNAse protection. Lines that confer the stabilized, stimulus-unresponsive half-life on the exogenous transcript would contain stabilization pathway mutations. Those that process the exogenous VEGF mRNA mRNA with a normal half-life would contain VEGF transcript mutations. Aim 3: Identify VEGF RNA sequence elements responsible for basal instability and for hypoxia- or TPA-induced message-stabilization. Determine whether these elements mediate constitutive stabilization in tumor cells. Regulatory RNA elements will be mapped using a panel of NIH3T3 clones expressing deleted forms the normal human VEGF mRNA. Inherent stability and hypoxia- or TPA-induced stability will be measured. Elements will be tested for transferable function in rabbit beta-globin fusion transcripts. Altered function of individual elements in tumor cell lines will be examined in the context of chimeric transcripts. Aim 4: Begin to identify, purify, clone, sequence, and characterize an RNA-binding protein(s) that mediates basal instability and hypoxia-induced upregulation of VEGF in normal cells, and constitutive upregulation in some solid-tumors. These proteins represent potential targets for anti-angiogenic therapy. Their identification and purification will be based on their binding to 32P-labeled RNA probes in nondenaturing gels and their lack of binding to mutant versions of specific RNA elements that were demonstrated to abrogate function in vivo (in Aim 3).

DESCRIPTION (provided by applicant): This is a revised renewal for the Molecular Pathology of Cancer Training Grant within the Molecular Pathology Graduate Program (MPGP) at UCSD School of Medicine. The purpose of this grant is to train predoctoral students to carry out cancer research based on a comprehensive knowledge of the molecular and cellular biology of cancer cells, of normal histology and cancer pathology, and of current cancer treatment. The training grant faculty are within the Cancer Biology Group of the MPGP, are located at UCSD School of Medicine, The Burnham Institute, and The Scripps Research Institute. The faculty are leaders in their fields of cancer research, which cover most topics in cancer cell biology, including alterations in signal transduction pathways involving protein kinases and phosphatases, mutations and/or altered expression of proto-oncogenes and tumor suppressor genes, the role of extracellular matrix and cell surface carbohydrates in cell adhesion and metastasis, genetic instability and chromosomal translocations, apoptosis, differentiation arrest in leukemia, the immune response to cancer cells, drug development, and cancer therapy. The Cancer Biology curriculum requires classes in The Molecular Pathology of Cancer, Mouse Models for Human Disease, Advanced Cancer Pathology, and two advanced electives in Human Genetics, Developmental Biology, Immunology, Biochemistry, Structural Biology, or Molecular Modeling (selected according to trainees research area). Students select a second Molecular Pathology class in human disease (Neurologic & Muscle disease, Cardiovascular development & disease, or Microbial Pathogenesis) to broaden their understanding of human disease. Trainees rotate through 3 laboratories prior to selecting their thesis advisor. In year 2, trainees take the Minor Proposition examination, which teaches them how to write an NIH-style grant, and evaluates their overall academic progress. To understand current cancer treatment and the need for molecular therapeutics, trainees attend a clinical conference, choosing among Tumor board, Breast Pathology, OB/GYN tumor board, and Hematology/oncology. They present annually at our MPGP Spring Research Retreat, at the annual Salk Institute meetings on Oncogenes or The Cell Cycle, and at Cancer Journal Club. They invite special seminar speakers annually. This training in molecular pathology of cancer combined with a basic understanding of medical oncology and cancer therapy is unique at UCSD.

DESCRIPTION (provided by applicant): HoxA9 and Meis1 cooperate to cause acute myeloid leukemia (AML), and are downstream effectors of a group of human AML oncoproteins, including MLL fusions. HoxA9 controls a self-renewal/differentiation switch and Meis1 confers leukemic potential, which correlates with its ability to activate transcription of stemness genes including Sox4, FLT3, and CD34. The biochemical and genetic mechanisms by which HoxA9 enforces self- renewal and by which Meis1 confers leukemic potential are not as yet understood. In this proposal, we use novel cell systems derived in our laboratory to address the key biochemical and genetic questions underlying myeloid leukemogenesis by HoxA9 and Meis1. In Aim 1, we focus on how apical oncoproteins activate transcription of Meis1. We will modulate Meis1 transcription using wild-type and conditional forms of MLL-ENL and NUP98-NSD1, a novel histone H3 Lysine 36 methyltransferase AML oncoprotein we recently discovered that coactivates transcription of Meis1 and HoxA9. Aim 2: In three different cell models, we discovered that the stem cell gene Sox4 is co-activated by HoxA9 and Meis1. Sox4 induces AML in mice and is activated in many other forms of human cancer. We hypothesize that Sox4 is a downstream effecter of HoxA9/Meis1 leukemogenesis. In Aim 2, we will characterize the Sox4 promoter, focusing on how HoxA9 and Meis1 cooperate with each other and other cofactors to activate Sox4 transcription. We hypothesize that HoxA9 and Meis1 catalyze distinct epigenetic modifications that cooperate functionally in activation. We will also deduce the requirement of Sox4 in HoxA9/Meis1 leukemogenesis, identify its genetic targets, and begin to pursue its cofactor function in establishing the AML phenotype. Aim 3: Terminal differentiation genes are strongly repressed by HoxA9, yet they are robustly activated when conditional forms of HoxA9 are switched off. In Aim 3, we will focus on how HoxA9 maintains active repression of terminal differentiation genes. For Aims 1-3, we use a novel technology called Chromatin Immunoprecipitation DNA Selection Ligation (ChIP-DSL), which will produce a high-resolution map of the histone marks of activation and repression, and of the binding sites for oncoproteins, transcription factors, and polymerases over 20 to 40 kbp of genomic DNA surrounding the transcriptional initiation site of each promoter. Lentiviral reporter vectors will be used to locate minimal promoters within responsive regions indicated by ChIP-DSL, and classical molecular and biochemical approaches will be taken to identify the activation mechanisms. Aim 4: We have already defined an N-terminal domain in HoxA9 and a C-terminal domain in Meis1 that are both required to activate transcription of Sox4, FLT3, and CD34. Our hypothesis is that each domain recruits a different histone modifying activity that cooperates in gene activation. In Aim 4, we identify cofactors that bind these domains, verify their importance in gene activation and in leukemogenesis, and begin to investigate the epigenetic changes they catalyze. These activities and the biochemical surfaces that recruit these cofactors are novel drug targets in AML. Project Narrative: This proposal focuses on how HoxA9 and Meis1 cause acute myeloid leukemia (AML) in humans, a cancer of blood stem cells. We focus on 3 of the most important questions: 1) How do human cancer proteins activate expression of the Meis1 leukemia gene? 2) How do HoxA9 and Meis1 activate expression of the Sox4 leukemia gene? 3) How does HoxA9 prevent expression of maturation genes? Discovering mechanisms underlying these questions may lead to the development of new AML therapeutics.