Description of Research Expertise

Research Interests
Neoplastic transformation by the Myc oncoprotein; microRNAs; tumor microenvironment.

Key words: Myc, p53, microRNA, angiogenesis, colon cancer, B-lymphoma, interferons, infection.

Description of Research
My laboratory studies the mechanisms of neoplastic transformation by the Myc oncoprotein as well as host responses to malignant growth, in particular tumor surveillance based on anti-angiogenesis (suppression of blood vessel growth). Myc is a transcription factor which regulates a perplexingly large number of genes. In the recent years, the interactions between Myc and other nuclear proteins and between Myc and promoter elements in genomic DNA have been exhaustively characterized. Yet to what extent these interactions define the oncogenic potential of Myc in vivo is understood only fragmentarily. On the other hand, the use of model organisms has revealed potent effects of Myc on normal and pathological development, but many underlying molecular mechanisms remain to be elucidated. Our goal is to understand gene regulation by Myc in the context of tumor progression. Thus, our research is directed towards: 1. Identifying Myc target genes in cell types corresponding to naturally occurring tumors; and 2. Establishing the functional significance of these target genes for tumorigenesis in vivo.

Our effort began with the discovery that Myc overexpression leads to down-regulation of thrombospondin-1, a secreted glycoprotein (J Biol Chem, 1996). Since thrombospondin-1 was known to inhibit angiogenesis, we hypothesized that by reducing its levels, Myc can promote the ingrowth of blood vessels. At that point, the role of Myc in tumor angiogenesis was not recognized. To put our hypothesis to a test, we demonstrated that transient overexpression of Myc indeed confers upon rodent fibroblasts the angiogenic phenotype (Cell Growth Diff, 2000). The effects of Myc on neovascularization were independently observed by several Myc laboratories, and recently it has been shown that down-regulation of thrombospondin-1 by Myc underlies the pro-angiogenic activity of Ras, another potent oncoprotein. The question still remained whether Myc can trigger the angiogenic switch in genetically complex tumors.

MYC PROMOTES ANGIOGENESIS IN MURINE COLON CARCINOMAS BY A MICRORNA-MEDIATED MECHANISM. To assess the angiogenic effects of Myc in a bona fide tumor context, we established a new mouse model of colon cancer. In this model, primary colonocytes deficient in the p53 tumor suppressor are sequentially transformed by Ras and Myc oncoproteins. We discovered that Ras/p53-null cells were very weakly angiogenic on their own. This was surprising, since mutations in both Ras and p53 had been reported to promote vascularization. However, in our system microvascular densities and overall growth rates became robust only after overexpression of Myc. Interestingly, the pro-angiogenic effects of Myc correlated with down-regulation of several proteins with thrombospondin-1 type repeats (TSR), including thrombospondin-1 itself (Tsp1) and connective tissue growth factor (CTGF). Previously we demonstrated that rather than affecting the thrombospondin-1 promoter, Myc decreases Tsp1 mRNA half-life (Nucl Acids Res, 2000). More recently, microRNAs have emerged as important regulators of mRNA stability, and at least one microRNA cluster (miR-17-92) is directly activated by Myc. Provocatively, thrombospondin-1 and CTGF are predicted targets for repression by miR-17-92. Consistent with this prediction, miR-17-92 knock-down with antisense 2'-O-methyl oligoribonucleotides partly restored Tsp1 and CTGF expression, and conversely, transduction of Ras-only cells with a miR-17-92-encoding retrovirus reduced Tsp1 and CTGF levels. Importantly, miR-17-92-transduced cells formed larger, better perfused tumors. These findings establish a role for microRNAs in non-cell-autonomous Myc-induced tumor phenotypes (Nature Genet, 2006; Nature Genet, manuscript under invited revision). Other TSR proteins, such as clusterin (a.k.a. ApoJ), also appear to be bona fide miR-17-92 targets. Importantly, down-regulation of clusterin by Myc boosts both proliferation and neovascularization of primary colonocytes. In addition, clusterin attenuates neoplastic transformation of epithelial cells during skin carcinogenesis (Cancer Res, 2004). These findings demonstrate how Myc shapes the angiogenic phenotype of solid tumors. However, recent research from Stanford University has demonstrated that the Myc->thrombospondin axis is functional in hematopoietic neoplasms as well. Thus, the role of Myc in lymphomagenesis is the second subject of our research efforts.

MYC REGULATES B-CELL MARKERS OF THERAPEUTIC IMPORTANCE. To determine the contribution of Myc to hematological malignancies, we developed another non-transgenic mouse model based on transduction of p53-null bone marrow cells with Myc-encoding retroviruses. We demonstrated that overexpression of Myc combined with inactivation of p53 suffices for B-lymphomagenesis (Oncogene, 2002). To gain a better understanding of the role of Myc in lymphomagenesis, we used a conditional mutant of Myc (MycER), which requires the presence of a synthetic estrogen (4-hydroxytamoxifen; 4OHT) for its activity. MycER/p53-null lymphomas were generated, and the role of Myc in tumor sustenance was studied using 4OHT deprivation. We discovered that inactivation of Myc does not cause overt tumor regression, as observed in one-hit transgenic systems. Instead, it merely suppresses cell cycle progression, leading to stasis and eventual relapse. However, the hallmark of surviving MycOFF cells is up-regulation of B-cell receptor (BCR) components as well as the interleukin-10 receptor and CD20, two well-known therapeutic targets. Thus, targeting Myc, while moderately effective on its own, shapes the phenotype of quiescent neoplastic cells and sensitizes them to other molecular therapies (Cancer Res, 2005; Ann N Y Acad Sci, 2005).

This experimental system proved to be very useful in the analysis of other Myc targets first identified in cultured cells. Already it has spawned numerous collaborations and joint publications with investigators interested in gene regulation by Myc, for example Chi Dang (Johns Hopkins University - Mol Cell Biol, 2006), Steven McMahon (Wistar Institute - Proc Natl Acad Sci USA, 2005) and Wafik el-Deiry (PENN - Mol Cell Biol, 2004). In these three studies, the MycER system allowed us to establish the role of Myc as a major regulator of transferrin receptor 1, metastasis-associated protein 1, and TRAIL-induced apoptosis inhibitor FLIP, respectively.

In parallel, we determined that some Myc-transformed cells possess dual B-myeloid potential and differentiate into macrophage-like cells following spontaneous down-regulation of the Pax5 transcription factor (Blood, 2003; Exp Cell Res, 2007). The propensity of some of our cell lines to silence Pax5 afforded a unique opportunity to study the role of Pax5 in B-cell differentiation. One such study, co-authored by Kathryn Calame's (Columbia), David Schatz's (HHMI-Yale) and our laboratories, was published several years ago (Nature Immunol, 2004). In that paper, we established that Pax5 controls commitment to the B cell lineage via the loss of histone 3 methylation in the VH immunoglobulin locus. A more recent study performed in collaboration with Dr. Michael Atchison (PENN) revealed that under the same conditions histone modifications at the Ig locus remain largely unchanged (J Immunol, 2006).

Interestingly, Pax5 appears to be an oncogene in its own right and cooperates with Myc in many non-Hodgkin lymphomas. Yet at the molecular and cellular levels, the contribution of Pax5 to neoplastic growth remains undeciphered. When we reconstituted our Pax5-deficient cells with Pax5ER, we observed greatly enhanced neoplastic growth. Expression profiling revealed that Pax5 is required to sustain expression of several crucial components of BCR, at least in the presence of Myc which attenuates BCR signaling (see above). The list of Pax5 activated genes included CD79a, a protein with the immunoreceptor tyrosine-based activation motif (ITAM). In contrast, expression of two known ITAM antagonists, CD22 and PIR-B, was suppressed. The key role of BCR/ITAM signaling in Pax5-dependent lymphomagenesis has been corroborated in experiments with constitutively active ITAM, forced expression of CD22, and a pharmacological inhibitor of an ITAM-associated tyrosine kinase Syk (J Clin Inv, in press). Thus, the interplay between Myc and Pax5 determines levels of BCR signaling and renders B-lymphoma cells sensitive to genetic and pharmacological inhibitors of this pathway.

MYC AND THERAPEUTIC APOPTOSIS. On the other hand, Myc is known to help maintain functional levels of the p53 tumor suppressor, at least in part via blocking its Mdm2-dependent ubiquitination. To assess the potential for therapeutic apoptosis in Myc-overexpressing lymphomas, we transduced Myc into hematopoietic cells with two knock-in alleles encoding the p53ER fusion. In the resultant tumors, 4OHT can trigger wide-spread apoptosis and overt tumor regression even in the absence of DNA-damaging agents. Interestingly, eventual re-growth of cell surviving p53-dependent apoptosis can be delayed by blocking autophagy (J Clin Invest, 2007). Yet in tumors with low level of Myc and high levels of Mdm2 even initial responses to p53 are blunted. However, co-treatment with proteasome inhibitors fully restores therapeutic effects in vivo. Similarly, human Burkitt's lymphomas with wild-type p53 and overexpression of Hdm2 are highly sensitive to proteasome inhibitors, unless p53 levels are reduced using the HPV-E6 ubiquitin ligase. Therefore, proteasome inhibitors could be highly effective as a monotherapy against Myc-induced lymphomas, with no need for adjuvant chemo- or radiation therapy. On the other hand, their efficacy is crucially dependent on the wild-type p53 status of the tumor, placing important restrictions on patient selection (Blood, 2007).

MYC TARGETS AND ANTI-TUMOR SURVEILLANCE. Our studies on Myc-induced lymphomas led to the conclusion that Myc profoundly down-regulates responses to several tumor-suppressive cytokines, including interferon gamma (IFNg). This prompted us to investigate the mechanisms underlying anti-neoplastic effects of IFN and to re-assess its therapeutic potential. We discovered that production of retrovirally encoded (Cancer Letters, 2001) or infection-induced interferon gamma strongly inhibits tumor angiogenesis and/or neoplastic growth (Cancer Biology & Therapy, 2003). Importantly, it is this inhibition of angiogenesis, not bystander immunity, that underlies well-documented resistance to tumors during infection. This became apparent when we observed tumor resistance in acutely infected mice lacking major cytotoxic responses, for instance T- and tumoricidal NK-cells (Journal of Immunology, 2001). Our discovery represented a major development in the field of tumor surveillance and was featured in numerous commentaries (LANCET, DRUG DISCOVERY TODAY, SCIENTIFICAMERICAN.COM, etc). One provocative corollary of our work is that anti-tumor vaccines should be evaluated not only for the ability to elicit cytotoxic immunity but also for their anti-angiogenic effects.

Rotation Projects for 2007-2008

1. Regulation of the thrombospondin-1 and related genes by Myc via a microRNA-based mechanism. The main goal of this project is to determine whether neoplastic growth by Myc-transformed colonocytes could be suppressed via enforced expression of these proteins and whether tumor suppression is due to cell-autonomous (e.g., cell cycle progression) or non-cell-autonomous (e.g., angiogenesis) processes.

2. The interplay between Myc, p53, and Pax5 in B-lymphomagenesis. We are interested in validating these genes as therapeutic targets. Our experimental data indicate that inactivation of Myc and Pax5 affects primarily cell cycle progression and B-cell differentiation, while re-activation of p53 results in immediate and wide-spread apoptosis and tumor regression.

3. Anti-angiogenic and tumoristatic properties of murine pro-inflammatory cytokines. We have already established that sustained production of interferon gamma blocks tumor growth, as does the deficiency in interleukin-10. We are now investigating the role of their downstream effectors, STAT1 and STAT3, in tumor cell proliferation, survival, and neovascularization.

Lab personnel:
Michael Dews, PhD, Senior Research Investigator
Elaine Chung, PhD, Postdoctoral Researcher
Prema Sundaram, PhD, Postdoctoral Researcher
Jason Jung, PhD student (Spring '08 rotation)
Waihay Wong, MD/PhD student (Spring '08 rotation)