Irving Weissman, MD
Dr. Weissman’s laboratory is working on identifying and characterizing the progression of discrete changes, genetic and epigenetic, that leads to the generation of cancer stem cells (CSCs) from a variety of blood and solid tissue cancers. They have found a single molecular event present in all cancers studied to date that protects them from macrophages of the innate immune system.
Myelodysplastic Syndrome
Multiple Myeloma
Acute Myelogenous Leukemia, Acute Lymphocytic Leukemia, and Lymphoma
Role of MicroRNAs in Normal and Malignant Hematopoiesis
Bladder Cancer
Brain Cancer
Melanoma
Epithelial Ovarian Cancer
Myelodysplastic Syndrome
Investigators: Wendy Pang, Ravindra Majeti, Beverly Mitchell, Peter Greenberg, Jason Gotlib, Tena Cherry, and Irving Weissman
This group has proceeded with their work on myelodysplastic syndrome (MDS), a group of hematopoietic disorders characterized by ineffective and dysplastic myelopoiesis. Clinically, patients with MDS suffer from cytopenias and an increased risk for progression to acute myelogenous leukemia (AML).
- In 2010, these investigators continued to characterize the differential expression of cell-surface markers on MDS hematopoietic stem cells (HSC). They have identified several differentially expressed markers, including CD109, CD53, CD164, and ITGA4 that distinguish MDS HSC from normal HSC.
- They have also identified a surface marker profile they believe has the potential to be a useful, new diagnostic tool for identifying MDS in patients, and they have been validating this tool on a larger set of samples.
- In addition, they have generated knockdown constructs (ways to reduce the expression of one or more of an organism’s genes) of RPS6, one of several ribosomal proteins whose expression is significantly downregulated in MDS, and have found that decreasing expression of these genes in normal HSC can cause developmental and proliferative phenotypes similar to those observed in the marrow of MDS patients.
Multiple Myeloma
Investigators: Dongkyoon Kim, Bruno Medeiros, and Irving Weissman
In collaboration with Bruno Medeiros, MD, assistant professor of medicine (hematology), the laboratory is extending their work in multiple myeloma (MM), an incurable B-cell cancer. They have also investigated the coexistence of two tumor-initiating cell populations and their evolutionary hierarchy in a human myeloma cell line, RPMI 8226.
- In 2010 these researchers found that the human bone marrow microenvironment is required for the engraftment of most primary myeloma cells into mice and that CD138+/CD38high plasma cells enrich long-lived, tumor-initiating cells while CD19+CD38low/- B cells do not.
- They also investigated the heterogeneity and the tumor initiating cells (TIC) of a human myeloma line, RPMI 8226. RPMI 8226 was composed of cells that varied in proliferation ability, immunophenotype, and morphology. Fractionation and serial transplantation of CD45+ and CD45- cells into immunodeficient mice indicated that both populations successfully formed tumors. The scientists demonstrated that RPMI 8226 cells are composed of at least two distinct TIC subsets and that new CD45- TICs are derived from preexisting CD45+ TICs. They propose that the heterogeneity of myeloma cells is maintained by the differentiation and clonal evolution of myeloma TICs.
Acute Myelogenous Leukemia, Acute Lymphocytic Leukemia, and Lymphoma
Investigators: Mark Chao, Max Jan, Ravindra Majeti, Ash Alizadeh, Sidd Jaiswal, and Irving Weissman
With a focus on the expression of CD47 on leukemia stem cells from AML and B-cell non-Hodgkin lymphoma (NHL) and its prognostic value, this group has made additional progress in the investigation of human AML and NHL in three major areas.
- Having previously demonstrated the prospective separation of normal HSC and leukemia cells on the basis of differential CD47 expression in a single AML patient sample, researchers Max Jan, a doctoral student, and Drs. Majeti and Weissman report the prospective separation of normal HSC from lymphoid stem cells (LSC) in multiple AML patient samples on the basis of differential expression of a newly identified AML marker, T-cell immunoglobulin mucin-3 (TIM-3). They determined that normal bone marrow HSC lack TIM-3 expression, while LSC from multiple AML specimens express high levels of TIM-3.These and other data provide the rationale for developing a panel of cell-surface markers, including CD47 and TIM-3, that enable the robust prospective separation of HSC and LSC for application to autologous hematopoietic cell transplantation and minimal residual LSC monitoring in the treatment of AML.
- MD/PhD student Mark Chao and Drs. Majeti and Weissman showed that blockade of CD47 with an anti-CD47 monoclonal antibody enabled phagocytosis of cancer cells leading to in vivo tumor elimination. They identified calreticulin as a pro-phagocytic signal necessary for targeting by CD47 blockade. While calreticulin, an “eat me” signal, is expressed on all cancer cells tested, only a minority of normal tissue cells, including only a minority of HSC, express calreticulin and are susceptible to anti-CD47 blockade. These findings identify calreticulin as the dominant pro-phagocytic signal on several human cancers, provide an explanation for the selective targeting of tumor cells (while sparing normal cells) by an anti-CD47 antibody, and highlight a new role for pro- and anti-phagocytic signals in the immune evasion of cancer.
- These researchers have identified increased expression of CD47 on human NHL cells and determined that higher CD47 mRNA expression independently predicts lymphocytic leukemia and mantle cell lymphoma. After demonstrating the eradication of human NHL in a mouse model using only a monoclonal antibody therapy that combines rituximab with a blocking anti-CD47 antibody, analyses of immune-cell subset contributions demonstrated that macrophages, but not NK cells or complement, were necessary and sufficient to enable antibody synergy in vitro and in vivo.
- These investigators have found that both human T-cell acute lymphoblastic leukemia (ALL) and B-cell ALL, malignancies of both children and adults, express CD47. Transplants of ALL to immune-deficient mice can take hold and spread as they do in leukemia patients. After these cancers become established and spread, Dr. Chao and collaborators have shown that treatment with blocking anti-CD47 antibodies leads to leukemia elimination in most cases, opening this type of antibody treatment alone, or with synergizing antibodies, to therapeutic clinical trials.
- These investigators now seek to determine whether antibodies synergized through a new mechanism combining Fc receptor (FcR)-dependent and FcR-independent stimulation of phagocytosis might apply to these and other cancers.
Role of MicroRNAs in Normal and Malignant Hematopoiesis
Investigators: Lisa Ooi, Chris Park, and Irving Weissman
Lisa Ooi, a doctoral student, and Dr. Weissman have been studying the profiles of miRNAs (small RNAs that negatively regulate gene expression) for normal HSC and progenitors in the human and mouse hematopoietic systems, as well as in primary human AML stem cells and non-stem cell blasts. They identified mir-125b as a microRNA highly expressed in the purified hematopoietic stem cell population as compared to downstream progenitors. Ectopic expression of mir-125b expands the mouse hematopoietic stem cell compartment and immature lymphoid progenitors through an anti-apoptotic mechanism as well as enriching for the lymphoid-balanced HSC subset. Effects of 125b are associated with the development of lymphoproliferative disease marked by expansion of CD8+ T lymphocytes. In the future, the investigators aim to describe the molecular and cellular requirements for miR-125b induced transformation, as well as the function of other stem cell miRNAs, in regulating leukemia stem cell function in primary human leukemia samples.
Bladder Cancer
Investigators: Keith Chan, Jens-Peter Volkmer, Robert Chin, Debashis Sahoo, Stephen Willingham, Mark Chao, Inigo Espinosa, Harcharan Gill, Joseph Presti, Matt van de Rijn, Ying Pan, Joe Liao, Mark Gonzalgo, Eddy Leman, Linda Shortliffe, and Irving Weissman
In previous work, these researchers defined CD44+ as a marker enriching for tumorigenic populations in bladder cancer. They have also shown that bladder cancer cells and, even more importantly, tumor-initiating populations express the “don’t eat me signal” CD47 and that blocking CD47 with a monoclonal antibody (antibody: Ab; monoclonal antibody: mAb) enables phagocytosis of cancer cells in vitro.
- This group has shown that therapeutic mAb targeting of CD47 on bladder cancer cells in patient xenograft tumors can significantly reduce tumor growth in a dose-dependent manner and that it can block metastasis to the lungs and lymph nodes in a clinically relevant xenograft model. Based on these data, CD47 may be a very promising target for bladder cancer treatment. They have also developed an orthotopic bladder cancer xenograft mouse model to determine the effect of local anti-CD47 Ab treatment alone or in combination with the established local immunotherapy, Bacillus Calmette-Guérin, commonly referred to as BCG).
- In 2010, these scientists identified CD47 as a diagnostic target for bladder cancer. In patient bladders that had been surgically removed due to advanced disease, they could specifically identify cancer tissue after application of flurochrome-labled anti-CD47 Ab followed by fluorescence microscopy within the bladders. In addition, the researchers developed an orthotopic bladder cancer xenograft mouse model to investigate CD47 Ab as a diagnostic tool in vivo.
- The group has developed a mathematical approach, termed MiDReG (Mining Developmentally Regulated Genes), to prospectively identify a CSC evolutionary hierarchy within bladder cancer using data from publicly available gene-expression datasets. For bladder cancer, this involved first selecting genes whose expression correlated with patient survival. Next, a gene hierarchy was identified based on cytokeratin (KRT14, KRT5, KRT20) expression patterns and corresponding cell-surface markers (CD90, CD44, CD49f). The team validated the hierarchical relationship between these cell populations in vivo in xenograft models. Finally, they showed that these three different subtypes of bladder cancer are strongly correlated with patient survival using both the cytokeratins and corresponding surface markers. This study supports the existence of multiple bladder CSCs with strong associations to patient survival and hierarchical relationships between them. The markers defining these CSC populations in the three bladder cancer subtypes are strong, independent markers to predict patient survival using both univariate and multivariate analysis.
Brain Cancer
Investigators: Siddhartha Mitra, Jian Wang, Albert Wong, Michael Edwards, Samuel Cheshier, Griffith Harsh, and Irving Weissman
This group has two main areas of interest: identification of cell-surface antigens on CSCs, which can be used for antibody therapy specific for brain tumor stem cells, and assaying the efficacy of anti-CD47 antibody therapy on brain-tumor-bearing mice, as well as identifying new or existing therapeutics to synergize with its antitumor effect.
- The goal is to identify CSCs that give rise to high-grade brain tumors, such as glioblastoma multiforme (GBM), and pediatric brain tumors, such as medulloblastoma. Progress on this project has slowed during the past six months due to the departure of its primary scientists. Two new scientists have joined the project: Siddhartha Mitra, PhD, a former postdoctoral fellow in the Department of Neurosurgery at Stanford Medical Center, and Jian Wang, PhD, from the laboratory of Per Enger, PhD, at the University of Bergen, Norway, who have studied brain tumors and stromal cells of the central nervous system (CNS). Miranda Orr, at the McLaughlin Research Institute in Great Falls, Montana, will join the group next year after finishing her doctoral studies on CNS stem cell populations from embryonic stem cells. Led by Dr. Mitra, the team has carried out large-scale analyses of functional cell-surface markers on multiple patient samples and identified three, new, putative cell populations that could potentially give rise to tumors. Cells bearing these markers have been sub-fractionated and orthotopically implanted into immunocompromised mice to assay their tumor-initiating potential. The investigators are also screening an anti-GBM monoclonal antibody library they generated to identify additional GBM-specific and GBM stem-cell-specific antigens, which will be used to identify potentially new combinatorial therapeutic options.
- Previously, these collaborators detected expression of CD47 on the majority of primary human brain tumors examined. An early study showed that daily, high-dose infusion of mouse CD47-specific mAb resulted in little or no loss of normal CD47+ brain cells in vivo and little toxicity outside the CNS. Preliminary experiments pre-coating primary medulloblastoma cells from patient samples with anti-CD47 Ab show that the antibody is able to significantly decrease tumor formation in an orthotopic xenograft mouse model. These scientists have also observed that intra-peritoneal injection of anti-CD47 antibody can efficiently penetrate the blood-brain barrier and coat tumor cells, making it a reliable mode of delivery. Currently, they are treating tumor-bearing mice with anti-CD47 Ab to assay its efficacy on existing brain tumors. From studies with other tissues, they hypothesize that another tumor-associated signal is probably required to trigger effective phagocytosis of target cells. EGFRvIII, a splice variant of the epidermal growth factor (EGF) receptor, is a tumor-specific antigen whose independent targeting has widely been used for immunotherapy. In collaboration with the laboratory of Albert J. Wong, MD, professor of neurosurgery, they found that CD47 and EGRvIII antigens frequently co-localize to the same brain tumor cell. Targeting both antigens may have a synergistic effect, which will be tested by using the two antibodies separately and, eventually, by combining antigen specificity in recombinant bispecific antibodies.
Melanoma
Investigators: Alex Boiko, George Yang, Arash Alizadeh, Michael Longaker, Denise Johnson, and Irving Weissman
Over the past year, these researchers profiled numerous melanoma samples from a broad spectrum of melanoma sites and stages and reported that human melanoma tumor stem cells (MTSC) express neural crest nerve growth factor receptor CD271.
- Using tumor cells directly isolated from surgical patient samples, they found CD271 to be heterogeneously expressed in nine out of ten melanomas analyzed, comprising between 2.5 percent and approximately 41 percent of the total cell population.
- These scientists also showed that human melanoma CD271+ cells engrafted in 70 percent of transplants (mice lacking B, T, and NK cells) compared to 7 percent of CD271− cells. They also demonstrated that in these same mice, tumors derived from transplanted human CD271+ melanoma cells were capable of metastasis in vivo. They minimized the potential selection of the most malignant cells due to serial passaging by using cells directly isolated from human surgical samples and transplanted using a protocol designed to minimize this artifact.
- Interestingly, they found that CD271+ melanoma cells in most patients lacked expression of well-defined tumor antigens, such as TYR, MART1 and MAGE, in 86 percent, 69 percent, and 68 percent of melanoma patients, respectively, which helps to explain why T-cell therapies directed at these antigens usually result in only temporary tumor shrinkage.
- Now that they have isolated melanoma tumor stem cells, this group plans to determine critical genetic and/or epigenetic alterations that lead to emergence, self-renewal, and differentiation of MTSCs. They plan to use advanced techniques, including microarray hybridization, to look for changes in signaling pathways driving melanoma development and progression. In addition, they will employ different strategies of targeting MTSCs, either by blocking antibodies to CD271 or by testing synergy with CD47/CD44 antibodies, to test whether melanoma regression occurs in mice transplanted with patient tumors.
Epithelial Ovarian Cancer
Investigators: Stephen Willingham, Jens Volkmer, Justin Cohen, Robin Martin, and Humberto Contreras-Truillo, Terry Storm, Nelson Teng, and Irving Weissman
These investigators have continued to collect primary epithelial ovarian cancer (EOC) samples for characterization and passage into xenografts. This is especially important in a disease as heterogeneous as EOC.
- This group has demonstrated that CD47 (originally named ovarian antigen 3, OA3) expression contributes to ovarian cancer growth and progression by protecting ovarian cancer cells from being eaten and killed by local macrophages. They have evaluated more than twenty tumors obtained directly from ovarian cancer patients and observed CD47 expression on nearly every tumor cell. Additional gene expression analysis of more than 400 ovarian tumor samples revealed that CD47 expression levels predicted clinical outcomes. Specifically, patients with tumors expressing high levels of CD47 had significantly worse long-term survival compared to patients with tumors expressing low amounts of CD47. They generated animal models that recapitulate the symptoms and tumor growth features observed in human patients to explore anti-CD47 antibody treatment and have now demonstrated that treating mice with anti-CD47 antibodies can not only dramatically inhibit the growth of human ovarian tumors in the peritoneal cavity but can also potentially eliminate growing tumors as a result of directly killing and/or blocking the ability of CD47 to prevent phagocytosis. Interestingly, the binding of one specific anti-CD47 antibody has the potential to directly induce the death of some ovarian cancer cells. Thus, the group believes that antibodies targeted to CD47 will promote clearance of ovarian cancer cells through two distinct mechanisms: 1) by blocking the “don’t eat me” signal transmitted by CD47, thereby promoting phagocytosis and 2) by the direct induction of cell death.
- In order to evaluate the therapeutic efficacy of CD47 target antibodies or to identify ovarian cancer stem cells, the research team has had to adapt current technologies to enable the direct visualization of growing ovarian tumors. Some EOC grow as a peritoneal ascites in patients; transfer of these cancers to the peritoneal cavity of highly immunodeficient mice leads to tumor growth as an ascites. A major drawback of these models, however, is the inability to continuously monitor tumor growth without harming the animal. In addition, patient tumors routinely require three to twelve months to grow in mice. Thus, the results of individual experiments were previously unknown until the animals progressed to advanced stages of disease. To accelerate progress, the investigators inserted the gene for luciferase (a bioluminescent enzyme) into several patient tumor cell populations, thereby enabling both continuous monitoring of tumor burden as well as earlier detection. These advances will undoubtedly accelerate ovarian-cancer-related research in the coming year.
- The California Institute of Regenerative Medicine granted the Weissman/Majeti disease team sufficient support to prepare good manufacturing practice-quality mAbs for toxicity tests in non-human primate models (produced using sufficiently high quality methods to ensure safety, purity, and effectiveness for human use) in preparation of joint clinical trials of AML patients at Stanford and in the United Kingdom. Most excitingly, the Stanford scientists have also shown that suppression of the CD47 “don’t eat me signal” is an effective therapeutic strategy for inhibiting the growth and metastases of a variety of blood and solid tumors, including ovarian, breast, colon, and bladder cancer. It is hoped that preclinical data with anti CD47 alone or in combination with other antibodies and/or chemotherapy protocols will lead to an early clinical trial in EOC.
- The Weissman epithelial ovarian cancer researchers continue searching to identify ovarian CSCs. Previous efforts that focused on established CSC markers failed to identify a marker that consistently selected ovarian cancer stem cells. Using high-speed analyzers, the group has recently profiled more than 100 new cancer and stem cell markers on four different patient ovarian tumors. The group wanted to determine whether the marker profile changed in mouse xenografts, so they also profiled three of the same tumors after passage through these animal models. This analysis has yielded at least ten new candidate markers that can be reliably used to separate out subsets of cells from ovarian tumors. Based on these results and advances in imaging ovarian cancer growth (discussed below), the research team expects to accelerate their evaluation of cancer stem cell markers in the coming year, after primary work on the CD47 project is completed.