Conference Report Molecular profiling and modeling: designing new strategies for the treatment of leukemia, lymphoma, and multiple myeloma. journalArticle pharmacotherapy, molecular medicine,hematology/oncology, internal medicine <b>Kris Novak, PhD</b><p><b>Clinical Editor: Sara M Mariani, MD, PhD</b> <b>Kris Novak, PhD</b>, is Senior Editor, <i>Nature Reviews Cancer,</i> London, United Kingdom. <b>Sara M Mariani, MD, PhD</b>, is Site Editor and Program Director, Medscape Hematology-Oncology. <BR> Novak K. FASEB 2001 Conference on Hematological Malignancies. MedGenMed 3(4), 2001 [formerly published in Medscape Hematology-Oncology eJournal 4(4), 2001]. Available at: http://www.medscape.com/viewarticle/408469 Medscape Medscape General Medicine <SUP>[TM]</SUP> 08/29/2001 3 4 <H3>Hematological Tumor Profiling</H3><FONT SIZE="2"> Microarray technology is one of the most recent approaches used to compare patterns of gene expression in 2 different tissues, such as tumor vs normal tissue or one type of tumor vs another. The procedure involves isolation of messenger RNA from the 2 different cell types, which is then reverse transcribed into cDNA. These 2 cDNA samples are then labeled with different fluorophores, and hybridized to a "DNA chip" that contains thousands of known genes. Resulting fluorescent intensities are collected using a laser confocal fluorescent microscope, and ratio information is used to compare expression levels for each gene between the 2 cell types.<p> This "expression profiling" is valuable for leukemia researchers, who hope to classify and subclassify the different hematological malignancies based on molecular data, rather than clinical information only. New techniques to classify these types of cancers are badly needed, since, often, it is difficult to predict how patients will progress and which patients will respond to certain therapies. DNA chips may someday be used in the clinic to determine patients' prognoses and whether they will respond to a specific therapeutic approach.<p><h4>Lymphochips and Lymphomas</h4> Louis Staudt,^[1] of the National Cancer Institute, Bethesda, Maryland, is studying gene expression in lymphoid malignancies using the cDNA microarray he calls the "lymphochip," which is a chip containing lymphocyte-specific genes. The lymphochip contains 15,000 genes of unknown function and 14,000-15,000 genes known to be involved in lymphoid development, isolated from lymphoid cDNA libraries. Last year, he showed that this system could be used to subdivide diffuse large B-cell lymphoma (DLBCL) patients into 2 categories.^[2]<p> Lymphoma cells from 1 set of patients expressed a set of genes associated with germinal center B cells, the site of antigen-driven B cell switch recombination, V(D)J gene hypermutation, and the selection process that generates high-affinity CD38⁺ memory B cells. On the other hand, cells from another subgroup of patients had a gene expression profile typical of circulating activated B cells. Patients with germinal center B-like lymphoma responded well to chemotherapy, with a cure rate of 75%. Patients with activated B-like lymphoma, however, did not respond well to chemotherapy.<p> Staudt also presented unpublished data showing that genes whose transcription is activated by the transcription factor NF-kB pathway are constitutively upregulated in activated B-like DLBCL. NF-kB is inactivated by cellular protein called IkB. Staudt's group generated activated B-cell-like DLBCL cells that produce a constitutively active form of IkB, and showed that expression of this protein leads to the death of these lymphoma cells in culture. Thus, agents that prevent NF-kB transcriptional activation may be developed as therapeutics for this subset of DLBCLs.<p> Staudt and colleagues are also using expression profiling to study B-cell chronic lymphocytic leukemia (B-CLL). Preliminary studies have revealed a gene expression signature that is shared by all CLL cases and that distinguishes CLL from all other lymphoid malignancies, suggesting that all CLL cases arise from the same precursor cell type or share a common mechanism of transformation.<p><h4>Microarray Analysis of B-CLL</h4> Riccardo Dalla-Favera's group, at Columbia University, New York, NY, is also using microarray analysis to study B-CLL cells.^[3] They have been using gene expression profiles to quantify relatedness between B-CLL cells and normal B-cell subsets, and have identified a number of CLL-specific genes. These include the guanine-nucleotide-exchange factor EPAC, the TGF-beta1 induced gene BIGH3, and the zinc finger transcriptional repressor BCL-6. BCL-6 is required for germinal center B-cell differentiation, and has been previously shown to be mutated in 35% of lymphoma cases. The fact that they isolated this gene is a good indication that their analysis is accurate.<p> BCL-6 is known to repress transcription of the chemokine genes MIP-1 and IP-10, along with c-Myc and the cell cycle inhibitor p27kip1. Therefore, malignant transformation by BCL-6 involves inhibition of differentiation and enhanced proliferation. Dalla-Favera reported that in normal cells, antigen stimulation and signaling through the surface protein CD40 downregulates BCL-6 at the transcriptional level. This downregulation, however, does not occur in lymphoma cell lines. In culture, Ramos lymphoma cells that constitutively express BCL-6 did not undergo NF-kB-activated expression of the costimulatory glycoprotein B7.1, which is involved in antigen presentation to T cells.<p> Dalla-Favera believes that BCL-6 binds to DNA at the site where the transcriptional enhancer STAT6/NF-kB normally would, preventing transcription of STAT6 and NF-kB target genes. Dalla-Favera showed that BCL-6 binds to the promoter region of the gene encoding B7.1, the costimulatory molecule necessary for B-cell/CD4+ T-cell interactions. Mutation of this binding site relieved the BCL-6 mediated repression of B7.1 expression. In Dalla-Favera's model, B7.1 expression is required to allow B cells to interact with T cells and then exit the germinal center. Lymphoma cells are trapped at this stage in B cell development because their constitutive expression of BCL-6 prevents expression of B7.1.<p><h4>Gene Expression in T-cell Acute Lymphoblastic Leukemia (T-ALL)</h4> Thomas Look,^[4] of Dana Farber Cancer Institute, Boston, Massachusetts, is examining gene expression signatures of different cases of T-ALL. T-ALL represents a subset (10% to 15%) of childhood leukemias that arise from oncogenes activated by translocations of the antigen receptor genes. Little is known about the molecular pathogenesis of this thymic cancer, and no new drugs are on the horizon.<p> Using quantitative polymerase chain reaction, Look identified three different T-cell oncogenes (LYL1, HOX11, and TAL1) that are upregulated in leukemic T cells in the absence of chromosomal translocations. HOX11 activation was associated with a favorable prognosis in 8 patients, while expression of TAL1 or LYL1 was not.<p> Using microarray analysis, Look identified 3 distinct gene expression signatures that were indicative of leukemic arrest at specific stages of normal thymocyte development. LYL1+ leukemic cells arrested at the pre-T-cell stage, expressing the surface protein CD34 and antiapoptotic genes such as Bcl-2. TAL1+ leukemic cells arrested at the late cortical thymocyte stage, expressing antiapoptotic genes, the T-cell receptor alpha and beta chains, p56lck, CD6, and the interleukin 8 receptor. HOX11+ cells arrested at the early cortical stage of thymocyte development. This is the stage at which T cells undergo positive and negative selection in the thymus, so these cells did not express any antiapoptotic genes.<p> Look proposed that HOX11+ cells are already "primed to die," and are therefore more susceptible to chemotherapy. These results suggest that the ability of gene expression profiling to stratify patients into subgroups can help to predict their response to chemotherapy.<p><h4>Gene Profiling in Multiple Myeloma (MM)</h4> Bart Barlogie's group used gene expression profiling to attempt to molecularly classify MM.^[5] New ways to subclassify MM are desperately needed, as it is particularly difficult to predict the course of this disease. Some patients progress slowly, while the disease can be rapidly fatal in others. Barlogie and colleagues performed a DNA microarray expression analysis of 6800 genes, comparing expression profiles of plasma cells from 63 primary MMs, 4 MM cell lines, and cells from 10 healthy bone marrow and 8 healthy tonsil samples. They are now beginning to analyze the 220 genes whose expression was deregulated in MM, including genes involved in cell adhesion, cell cycle, drug resistance, signal transduction, and apoptosis.<p> They found 57 genes whose expression increased dramatically in certain subsets of MM, such as the fibroblast growth factor receptor-3, which was specifically upregulated in 100% of t(4;14)(p16;q34)-related MM cases. They are now trying to determine if the upregulation of any of these other genes is associated with specific translocations.<p> Barlogie and colleagues are also comparing gene expression in myelomas that carry chromosome 13 deletions vs those that do not. It is known that deletions of portions of chromosome 13 are associated with a more aggressive form of MM. About 28% of MM patients who do not carry a chromosome 13 deletion are in complete disease remission 5 years after diagnosis, while none of the patients carrying a chromosome 13 deletion is in complete remission at this time point.<p> Barlogie's group analyzed the expression pattern of 84 genes located on chromosome 13, searching for those genes whose expression changed dramatically in MM cells. They discovered a pattern of gene expression that correlated well with the deletion, including alterations in the expression of the genes named SAP18, RPL21, and IMOGEN38. The exact function of these gene products is not clear yet, and they are planning to study their potential role in the development of MM.<p></font><p><P><H3>Chronic Myeloid Leukemia (CML) and Acute Myeloid Leukemia (AML)</H3><FONT SIZE="2"> This year has been an exciting one for CML researchers, clinicians, and patients, as the kinase inhibitor STI571 was approved in record time for the treatment of this B-cell cancer. CML is caused by the Philadelphia chromosome translocation, which generates the activated tyrosine kinase fusion protein Bcr-Abl. STI571 inhibits kinase activity of Bcr-Abl, and was recently reported to induce complete remission in 98% of chronic-phase CML patients.^[6]<p><h4>Mechanisms of Resistance to STI571</h4> Charles Sawyers,^[7] of the University of California, Los Angeles, who was involved in the first clinical trials of STI571, discussed mutations found in patients who have become resistant to the drug. He reported that in addition to the effects seen in chronic-phase patients, response rates to STI571 therapy were high in blast crisis patients, despite the presence of multiple oncogenic abnormalities in addition to Bcr-Abl. However, relapses occur frequently in treated blast crisis patients. Through analysis of blood and bone marrow cells from these patients, Sawyers and colleagues have been studying mutations that cause STI571 resistance. Cells from some of the patients analyzed were found to possess mutations that restored Bcr-Abl function.<p> Sawyers' assay for Bcr-Abl function involved measuring phosphorylation levels of the kinase's substrate, CRKL. Levels of CRKL phosphorylation have been shown to increase in CML patients who possess the Philadelphia chromosome, and decrease after STI571 treatment. In 14 blast crisis patients who no longer responded to the drug, CRKL phosphorylation was restored. Three patients were found to have amplifications in the Bcr-Abl gene, with 5, 10, and 15 copies of the gene. Gene amplification may cause production of too much fusion protein for the drug to inhibit. Alternatively, 6 patients were found to have a mutation in the ATP binding pocket of the Abl kinase, the site where STI571 binds to the protein. In this case, the drug can no longer bind and inactivate the cancer-causing kinase.<p> Sawyers suggested that in view of the frequency of secondary mutations, blast crisis patients should receive STI571 plus standard chemotherapy or interferon.<p><h4>AML-1/ETO and Pathogenesis of AML</h4> Scott Hiebert,^[8] of Vanderbilt-Ingram Cancer Center, Vanderbilt University, Nashville, Tennessee, discussed pathogenic mechanisms of AML. About 12% of AML cases are associated with the chromosomal translocation t(8;21), which creates a fusion protein between the transcription factor AML-1 and the transcriptional repressor ETO (AML-1/ETO). Hiebert and colleagues searched for genes that would be repressed by the AML fusion proteins, and found that the tumor suppressor p19^ARF promoter contains 8 perfect AML1 bindings sites. P14^ARF is a cell cycle checkpoint protein that is involved in promoting cell cycle arrest.<p> Both AML and AML-1/ETO bind specifically to the AML-1 site in the p14^ARF promoter. AML-1 activates p14^ARF transcription, but AML-1/ETO, on the other hand, represses p14^ARF transcription in reporter assays and reduces endogenous p14^ARF expression in multiple cell types. In human AML samples containing the t(8;21) translocation, levels of p14^ARF are, on average, 10-fold lower when compared with other AMLs. Thus, AML-1/ETO may transform cells by downregulating expression of the tumor suppressor p14^ARF.<p><h4>Myeloid Transcription Factors in AML Cells</h4> Daniel Tenen,^[9] of the Harvard Institutes of Medicine, Harvard Medical School, Boston, Massachusetts, spoke about the myeloid transcription factors C/EBP-gamma, which play important roles in the development of the myeloid lineage. Hematopoiesis in C/EBP-gamma null mice is blocked at the earliest stage of granulocyte differentiation and these mice have myeloid blasts in the blood, similar to AML patients. Tenen reported that this transcription factor is inactivated in human AML patients.<p> A number of patients who do not carry the t(8;21) or t(15;17) translocations were found to have heterozygous mutations in C/EPBP-gamma. One subgroup of AML1-ETO-positive patients have 8-fold lower C/EBP-gamma expression levels than other subgroups of AML patients. C/EBP-gamma protein was undetectable in AML-ETO-positive AML patients' cells or cell lines. Conditional expression of AML-1/ETO in cells downregulated C/EBP-gamma expression, and resulted in granulocytic differentiation of AML1-ETO-positive hematopoietic cells. Tenen suggested that restoring C/EBP-gamma &#61472;expression may be a useful strategy to treat AML1-ETO-positive leukemias.<p></font><p><P><H3>Acute Promyelocytic Leukemias (APLs)</H3><FONT SIZE="2"> APLs make up about 10% of all AML cases. In APL, leukocyte precursors are trapped at the promyelocytic stage, and expand. APL patients respond to therapy with all-trans-retinoic acid (ATRA), which induces differentiation, growth arrest, and apoptosis of leukemic cells. However, many patients undergo relapse and become resistant to ATRA therapy, so new therapeutic approaches are needed.<p> Several different fusion proteins, created by chromosomal translocations, are associated with APL, including fusions between the transcription factors PML, PLZF, NPM, NUMA or STAT5b, and the retinoic acid receptor-alpha. APL researchers are trying to understand how these fusion proteins induce a block in myeloid differentiation and cause cancer.<p><h4>Histone Deacetylase (HDAC) in APL</h4> Pier Giuseppe Pelicci, of the European Institute of Oncology, Milan, Italy, discussed the role of HDAC in APL.^[10] One form of APL is caused by the chromosomal rearrangement t(15;17), which creates a fusion protein between RAR-alpha and PML, a phosphoprotein that functions as a transcription factor and tumor suppressor. The fusion protein oligomerizes, recruits HDAC, and becomes a constitutively active transcriptional repressor.<p> Oligomerization and altered recruitment of HDACs are also responsible for transformation by the fusion protein AML-1/ETO, extending this mechanism to other forms of AML. These findings suggest that HDAC is a common target in myeloid leukemias, and that HDAC inhibitors may be used to treat patients with AML.<p><h4>Protein Acetylation in APL</h4> Arthur Zelent, of the Leukaemia Research Fund Centre at the Institute of Cancer Research, Chester Beatty Laboratories, London, United Kingdom, spoke about the role of protein acetylation in acute promyelocytic leukemia (APL).^[11] Abnormal recruitment of the nuclear receptor corepressor (N-CoR) and HDAC by leukemia-associated fusion proteins plays a role in pathogenesis of APL. In APL, fusion of RAR to the transcriptional repressors PML or PLZF results in expression of chimeric oncoproteins that are defective in corepressor/coactivator exchange following ATRA binding.<p> PLZF, which is thought to regulate growth and differentiation of hematopoietic cells, has been shown to recruit components of the N-CoR/HDAC complex in an ATRA-independent manner, accounting for the ATRA-resistance seen in forms of APL associated with the RAR/PLZF fusion protein.<p> Zelent reported that PLZF can associate with other transcriptional coactivators, such as p300, SRC, and pCAF, which possess histone acetylase activities. Acetylation of PLZF by p300 appears to stimulate its ability to repress transcription and cause deacetylation of histone H4 in vivo. Furthermore, PLZF can be deacetylated by HDAC3, relieving its ability to repress transcription. Zelent also described the discovery of a new class II HDAC that is involved in this process.<p> These results suggest that HDAC-dependent transcriptional repressors can be regulated through a cycle of acetylation/deacetylation, and that HDAC and histone acetylases play an important role in regulating gene expression during leukemogenesis. Accordingly, Zelent suggested that HDAC inhibitors may be used to treat APL.<p><h4>Arsenic Trioxide and Therapy of APL</h4> Hughes deThe, of the Hospital St. Louis, Paris, France, discussed mechanisms underlying the therapeutic use of arsenic trioxide, which has been shown to induce complete remission in some patients with APL.^[12] As2O3 is believed to function by inducing apoptosis of leukemia cells, as shown ex vivo. The drug has also been shown to induce differentiation and apoptosis of leukemic cells in vivo with few side effects.<p> DeThe showed that in APL, the PML/RAR-alpha fusion protein disrupts formation of PML nuclear bodies and nuclear matrix structures that are believed to be involved in regulating transcription. Successful APL treatments, such as retinoic acid or As2O3, have been shown to restore nuclear body formation. PML is modified by the ubiquitin-related peptide SUMO, a process enhanced by As2O3 and thought to target PML to the nuclear matrix.<p> DeThe demonstrated that As2O3 triggers the proteasome-dependent degradation of PML and PML/RAR-alpha. Mutation of PML's sumolation site revealed that PML sumolation is not required for its As2O3-induced matrix targeting or formation of early nuclear bodies, but it is required for the formation of mature nuclear bodies. Thus, PML sumolation promotes its own catabolism, suggesting that mature nuclear bodies could be sites of intranuclear proteolysis.<p></font><p><P><H3>T and B Leukemias: Mouse Models</H3><FONT SIZE="2"> New animal models are greatly needed to study hematological malignancies, as they provide insight about disease pathogenesis and allow for testing of new therapeutics. Several scientists at the meeting have developed promising new mouse models that are shedding more light on cancer mechanisms.<p><h4>The Oncogene TCL1 in T-cell Lymphocytic Leukemia (T-CLL)</h4> Carlo Croce's group at the Thomas Jefferson Medical College, Philadelphia, Pennsylvania, is developing mouse models to study the role of the oncogene TCL1 in T-CLL, the most common of mature T-cell malignancies.^[13] T-CLL has been associated with recurrent reciprocal translocations in chromosome 14. These translocations activate transcription of TCL1, leading to its overexpression in adult T-CLL patients. TCL1 is expressed in pre-B cells and in immature thymocytes, but not in more differentiated CD4+ and CD8+ subpopulations, suggesting a potential role in early lymphoid development.<p> Croce reported that mice overexpressing TCL1 develop a hemopathy similar to human T-CLL. Alternatively, when TCL is placed under control of an immunoglobulin promoter, mice develop B-cell leukemia. Conversely, TCL1 knock-out mice have a reduced number of B cells, suggesting that TCL1 is required for development of the lymphoid cells in which it is expressed.<p> But what is TCL1's exact function in lymphoid development? TCL1 interacts with the oncogene Akt, activating its kinase activity, and also plays a role in its nuclear translocation. In the nucleus, Akt phosphorylates a transcription factor called Nur77, thus blocking its ability to activate transcription of proapoptotic genes. Therefore, according to studies in mouse models, TCL1 seems to promote cell proliferation and to prevent apoptosis of T-cell progenitors.<p><h4>BM-1, Downstream Mediator of E2a-Pbx1</h4> Michael Cleary, of Stanford University, Stanford, California, discussed the role of Bmi-1 as an essential downstream mediator for the oncogenic effects of E2a-Pbx1 in hematopoietic cells.^[14] Although many acute B-cell precursor leukemias contain inactivating deletion or mutations in the INK4A-ARF gene, there is also a subset of t(1;19) chromosomal translocations that lack mutations in this gene. This translocation is present in 20% of pre-B-cell leukemias and generates a chimeric transcription factor known as E2a-Pbx1, a fusion protein containing the strong transactivation domains of E2a and the DNA binding domain of Pbx1. Pbx is a Hox DNA binding cofactor.<p> The activated oncoprotein E2a-Pbx1 transforms several cell types in vitro and induces lymphoblastic lymphomas in transgenic mice. Cleary suggested that E2a-Pbx1 functions as a "rogue activator" whose regulatory properties differ from wild-type Pbx1 but remain dependent on Hox DNA-binding partners.<p> Little is known about the transcriptional pathways that are perturbed by E2a-Pbx1 or how they contribute to leukemogenesis. Cleary, therefore, used cDNA microarrays to find E2a-Pbx1-responsive genes. By fusing a metal responsive element to the E2a-Pbx1 gene, Cleary and colleagues were able to induce fusion protein expression in B-cell progenitors by addition of ZnSO₄. They compared expression patterns of induced vs uninduced cells by microarray analysis. E2a-Pbx1-expressing cells were found to upregulate the gene Bmi-1, along with a number of other genes. Expression of Bmi-1, a known transcriptional repressor of the INK4a/ARF tumor suppressor, was upregulated within 24 hours of fusion protein induction.<p> Cleary showed that in B-cell progenitor cells, conditional expression of E2a-Pbx1 induced expression of BM-1, which suppressed expression of p16^Ink4A and p14^ARF, resulting in S-phase cell cycle entry. E2a-Pbx1 induction of Bm-1 also postponed replicative senescence of human diploid fibroblasts. Restoring p16INK4A activity to these cells reversed this effect. Bmi-1-deficient hematopoietic progenitors and mouse embryo fibroblasts are resistant to transformation by E2a-Pbx1. Thus, Bmi-1 is likely to be an important mediator of the translocation product E2a-Pbx1 in a subset of human lymphoid leukemias.<p><h4>Notch-Mediated Signals in Leukemic Cells</h4> Warren Pear, of the University of Pennsylvania, Philadelphia, Pennsylvania, discussed the role of Notch signaling in leukemia cells.^[15] A subset of T-ALLs contains a t(7;9) chromosomal translocation, which juxtaposes the TCR locus with that of Notch1, resulting in aberrant expression of Notch1. Notch1 is an oncogene that causes development of T-ALL when expressed transgenically in mouse bone marrow. Pear and colleagues used this model to study the Notch1 signaling pathway involved in T-cell transformation. Structure/function analysis revealed that both the ankyrin repeat and the c-terminal transactivation domains are required for transcriptional activation and for leukemia induction, suggesting that Notch requires transcriptional cofactors to transform T cells.<p> Another question that Pear investigated was why activation of Notch1 in the bone marrow only results in T-cell leukemias, and not B-cell leukemia. Examination of the transgenic mice revealed that they did not possess B-cell precursors at any stage of differentiation. Pear proposed that, in his model, Notch1 is a developmental switch and its expression by lymphoid progenitors directs them all toward the T-cell differentiation pathway.<p> So how does Notch1 function in T-cell development? Unlike Ras or Lck, Notch is unable to induce differentiation of RAG2-/- hematopoietic progenitors to CD4⁺/CD8⁺ T cells. However, Notch1 is able to induce expansion of a very early T-cell progenitor (CD3⁺, CD25⁺, CD44^- cells) in mice. Thus, Notch1 is required only for the early stages of T-cell commitment. When activated Notch1 was introduced into bone marrow progenitor cells from RAG2-/- mice that expressed a transgenic T-cell receptor, mature CD4⁺/CD8⁺ T cells did develop and expand, and these mice developed a clonal T-cell leukemia.<p> These results suggest that Notch1 drives differentiation of a lymphoid progenitor toward a T-cell fate, and that subsequent T-cell receptor rearrangement allows T-cell expansion and maturation to the double positive stage.<p><h4>RUNX1-Mediated Transcription and Hematopoiesis</h4> Nancy Speck,^[16] of the Dartmouth Medical School, Hanover, New Hampshire, addressed the roles of the transcription factors Runx1 and core-binding factor (CBF) in hematopoiesis. RUNX1 (AML1) and CBFB are the most frequently mutated genes in human leukemia cells. RUNX1 null mice do not undergo hematopoiesis (although they do undergo erythropoiesis).<p> Hematopoietic stem cells first emerge from the yolk sac, vitelline and umbilical arteries, and the aorta/gonad/mesonephros region. Runx1 is expressed by mesenchymal and endothelial cells in all these sites, but at no other locations in the mouse embryo. Speck discussed how RUNX1-null bone marrow cells cannot reconstitute any amount of hematopoiesis in irradiated mice, thus demonstrating that Runx-1 is required for the earliest stages of hematopoiesis.<p></font><p><P> <ol><li>Staudt LM. Gene expression profiling of lymphoid malignancies reveals new cancer types and mechanisms of malignant transformation. Program and abstracts of the FASEB 2001 Conference on Hematological Malignancies; July 28-August 2, 2001; Snowmass, Colorado. <li>Alizadeh AA. Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature. 2000;403:503-511. <li>Dalla-Favera R. Microarray analysis of B cell chronic leukemia. Program and abstracts of the FASEB 2001 Conference on Hematological Malignancies; July 28-August 2, 2001; Snowmass, Colorado. <li>Look AT. Molecular pathogenesis of T-cell acute lymphoblastic leukemia. Program and abstracts of the FASEB 2001 Conference on Hematological Malignancies; July 28-August 2, 2001; Snowmass, Colorado. <li>Shaughnessy J, Barlogie B. Gene expression profiling in the molecular characterization of multiple myeloma. Program and abstracts of the FASEB 2001 Conference on Hematological Malignancies; July 28-August 2, 2001; Snowmass, Colorado. <li>Druker BJ. Activity of a specific inhibitor of the BCR-ABL tyrosine kinase in the blast crisis of chronic myeloid leukemia and acute lymphoblastic leukemia with the Philadelphia chromosome. N. Engl J Med. 2001;344:1084-1086. <li>Sawyers CL. Developing kinase inhibitors for leukemia treatment: CML. Program and abstracts of the FASEB 2001 Conference on Hematological Malignancies; July 28-August 2, 2001; Snowmass, Colorado. <li>Hiebert S. AML-1/ETO binds multiple co-repressors and class I histone deacetylases and represses the p14^ARF tumor suppressor, a mediator of the p53 oncogene checkpoint. Program and abstracts of the FASEB 2001 Conference on Hematological Malignancies; July 28-August 2, 2001; Snowmass, Colorado. <li>Tenen DG. Myeloid transcription factors, differentiation and leukemia. Program and abstracts of the FASEB 2001 Conference on Hematological Malignancies; July 28-August 2, 2001; Snowmass, Colorado. <li>Minucci S, Nervi C, LoCoco F, Pelicci PG. Histone deacetylases: a common molecular target for differentiation treatment of acute myeloid leukemias? Program and abstracts of the FASEB 2001 Conference on Hematological Malignancies; July 28-August 2, 2001; Snowmass, Colorado. <li>Guidez F, Petrie K, Ivins S, Hawe N, Howell L, Zelent A. Protein acetylation and acute promyelocytic leukemia. Program and abstracts of the FASEB 2001 Conference on Hematological Malignancies; July 28-August 2, 2001; Snowmass, Colorado. <li>Lallemand-Breitenbach V, Zhu J, Puvion F, et al. Role of PML sumolation in nuclear body formation and As2O3-induced PML or PML/RAR degradation. Program and abstracts of the FASEB 2001 Conference on Hematological Malignancies; July 28-August 2, 2001; Snowmass, Colorado. <li>Croce C. Role of TCL1 in T cell prolymphocytic leukemia. Program and abstracts of the FASEB 2001 Conference on Hematological Malignancies; July 28-August 2, 2001; Snowmass, Colorado. <li>Chandra S, Smith K, Lingbeek M, et al. Bmi-1 is an essential downstream mediator for the oncogenic effects of E2a-Pbx1 in hematopoietic cells. Program and abstracts of the FASEB 2001 Conference on Hematological Malignancies; July 28-August 2, 2001; Snowmass, Colorado. <li>Izon D, Allman D, He J, et al. Notch signaling in lymphoid cell fate decisions and leukemia. Program and abstracts of the FASEB 2001 Conference on Hematological Malignancies; July 28-August 2, 2001; Snowmass, Colorado. <li>Speck N. Runx-1-CBFb is required at multiple stages of hematopoiesis. Program and abstracts of the FASEB 2001 Conference on Hematological Malignancies; July 28-August 2, 2001; Snowmass, Colorado. </ol>