Annual Forum 2006 – Stem Cells

John E. Dick, Ph.D., University of Toronto, Ontario, Canada
David Scadden, MD, Massachusetts General Hospital, Boston, MA

Philip Beachy, Ph.D., Stanford University, Stanford, CA
Ann Chambers, Ph.D., London Health Sciences Centre, London, Ontario, Canada
Michael F. Clarke, MD, Stanford University School of Medicine, Palo Alto, CA
Hans Clevers, MD, Ph.D., Uppsalalaan - Hubrecht Laboratory, Utrecht, the Netherlands
R. Keith Humphries, British Columbia Cancer Agency, Vancouver, British Columbia, Canada
Craig T. Jordan, Ph.D., University of Rochester School of Medicine, Rochester, NY
Haifan Lin, Yale University School of Medicine, New Haven, CT
Luis F. Parada, Ph.D., University of Texas Southwestern Medical Center, Dallas, TX
Professor Chris Potten, EpiStem, Ltd., Manchester, UK
Derek van der Kooy, Ph.D., University of Toronto, Toronto, Ontario, Canada
Amy Wagers, Ph.D., Joslin Diabetes Center, Boston, MA
Irv Weissman, MD, Stanford University School of Medicine, Stanford, CA

Layout and Goals

  1. Stem Cell Properties
  2. Self Renewal
  3. Cancer Stem Cells & Environmental Control
  4. Therapy & Metastasis

The focus of this meeting will be on tumor stem cells; how they arise and how they differ from the progeny that originate from these stem cells. Until 1998, stem cell biology was a field of narrow interest to developmental biologists with a single point of contact to adult human medicine: bone marrow transplantation. The first human embryonic stem cell lines were reported that year and a virtual revolution was begun. Over the same time interval, an explosion of information about adult stem cells in many tissues has emerged albeit with less of a fanfare and more realistic expectations. At least two consequences of the rapid developments in these areas are relevant to the cancer problem. First, is a greater understanding of the normal stem cells that exist in tissues and how these cells both regenerate and differentiate into normal tissue counterparts. If one understands how normal cells behave there is more of a chance that we will understand the malignant process. This should in turn help to define approaches to destroying cancer cells in the body. Second, the relationship of a normal stem cell to its environment is viewed as critical for its regulated growth and its ultimate survival. Studies in this field may shed light into the problem of metastasis and again identify new approach to treatment. Bringing together experts in the areas of stem cell biology, should enable us to hold a Forum on cancer stem cells and metastasis; topics that are highly relevant today to defining what we all seek; a cure for cancer.

Outcome Report
The 2006 Forbeck Forum focused on a question that has captured the attention of cancer scientists and cancer research funding bodies around the world. Does every cell in a tumor, whether liquid (leukemia) or solid, have equal ability to sustain cancer growth or are some cells within the tumor more potent than others? The answer to this question has the potential to alter research approaches away from studying the cellular and molecular properties of entire tumor tissue towards focusing on the tumor-initiating cells or so called cancer stem cells (CSC). There was much discussion at the Forum on the best nomenclature for these cells. Evidence is emerging that for many tumors, they are organized as cellular hierarchies that are sustained by stem-like cells in much the same way as normal organs. To gain a clearer picture of CSC and their similarities and differences to normal stem cells, the 2006 Forum assembled 12 leading stem cell scientists whose interests ranged from stem cell biology of lower organisms like Drosophila to blood and intestinal stem cells of the mouse, to normal human hematopoietic stem cells (HSC) and leukemic stem cells (LSC). The unifying theme of the Forum was that progress to identify and characterize CSC from different tumors and to understand their importance in cancer will only come when we understand how normal stem cells for each organ actually work. One field of study, normal or neoplastic, informs the other. Indeed by understanding the genetic and epigenetic programs that govern normal stem cells, we can begin to understand how the neoplastic process subverts normal stem and progenitor cells. This Forum showed once again the value in bringing together investigators from different stem cell areas that often do not talk to each other and having them focus their attention on one question. New research directions were formulated, collaborative research projects were developed, and ultimately a strategy for progress in this promising field was stimulated in each of the participants. As with many areas of biology our understanding of regulatory mechanisms that stem cells use comes from lower organisms.

Philip Beachy described studies of a central pathway in pattern formation in Drosophila, the hedgehog (Hh) pathway and how this pathway regulates stem cells from several different tissues and how this pathway becomes subverted in cancer. For example, Hh activation in the skin leads to basal cell carcinoma. Yet pathway blockage in normal tissues does not result in abnormalities unless tissue injury ensues. He described how injury drives otherwise quiescent stem cells to be activated becoming migratory and proliferative. If pathway activity is sustained because of injury or inflammatory processes, additional epigenetic and genetic changes can arise resulting in the generation of abnormal CSC. He demonstrated that even in non-epithelial tissues, such as multiple myeloma (MM), the MM-initiating cell (MM-IC) fractions had higher Hh pathway activation than the larger non-MM-IC fractions. This knowledge can be harnessed by developing therapies that block Hh activated CSC.

Hans Clevers described how his long term studies of the genes that play a central role in the growth and differentiation of the cell types that make up the small and large intestine are revealing the genetic signature for each developmental stage from an intestinal stem cell, to the transit amplifying (TA) progenitors, to the mature lineages. Deregulation of many of these genes results in abnormal growth and neoplastic growth. He showed the role of the WNT pathway on the proliferation of TA progenitors, the Notch pathway in lineage determination, and the tissue specific pattern of function of effector genes such as Tcf4. Finally, he provided evidence of an intestinal stem cell in an unexpected location amongst the Paneth cells of the small intestine or the bottom of the crypt of the large intestine. These cells express a putative stem cell marker, GPR49. These studies provide the first glimpse of this new intestinal stem cell and how it may be involved in early steps of adenoma formation, the first step of the neoplastic process. Given the expression in normal stem cells and a small subset of neoplastic cells, this new marker may be pointing to a new marker for a colon CSC showing the power of combining normal stem cell and cancer stem cell studies.

The Forbeck Scholar Joseph Wu talked about his work aimed at controlling human embryonic stem cell (hES) development into mesodermal lineages and ultimately into cardiac mesoderm under controlled conditions. He developed a genetic reporter construct of cardiac differentiation that enabled multiparameter imaging by flow cytometry, bioluminescence, and functional imaging by PET. Collectively imaging techniques such as this will have important uses in all of stem cell biology. Evidence is emerging that epigenetic factors play a large role in regulating stem cell function. Certain repressors expressed in stem cells prevent the activation of differentiation programs. However how these epigenetic factors function in stem cells is poorly understood. Haifan Lin again demonstrated how discovery of fundamental control mechanisms of stem cells from lower organisms can lead to new insights of mammalian stem cell biology. He described the pathway of discovery from identification of the role that the gene PIWI plays in germ stem cells in Drosophila to its involvement in epigenetic silencing, in the process linking non-coding RNA such as RNAi and miRNA to epigenetic silencing. He showed that some proteins that interact with PIWI are canonical epigenetic factors that together to create a functional unit that regulates chromatin. PIWI and some PIWI interacting proteins such as HP1 regulate self renewal, a biological property that is central to the stem cell state. Moreover since PIWI genes are also involved in RNAi processing and function, this new work links non-coding RNA to chromatin regulation in addition to more well known functions of RNAi in post-translational modification. Interestingly the human version of PIWI, called HIWI is dysregulated in human germ cell tumors and GI cancer. Silencing of HIWI in human cancer cell lines results in reduced growth; pointing to central importance in the neoplastic process. Since stem cells are long lived and able to produce large numbers of progeny, mechanisms must be in place to guard against DNA damage within the stem cells and to prevent neoplastic transformation. One such mechanism was proposed some 40 years ago by Cairns, that suggested since many stem cells must divide asymmetrically to yield one daughter stem cell and one committed daughter cell, the original DNA template of the parental stem cell segregated to the daughter stem cell only. Since this template is not copied it is much less susceptible to mutation.

Derek van der Kooy described a series of experiments to determine if this model held for neural stem cells (NSC). He utilized as an assay for NSC in vitro sphere formation and marked primordial and newly replicated DNA strands with BrdU and 3H-TdR. He showed evidence of unequal strand segregation into NSC daughters. However, he also described the difficulties in showing that the resulting daughters were true NSC and that every NSC segregated DNA in this fashion. Therefore he turned to a more robust stem cell system of the germ stem cells of the fly ovary where cell position denotes stem cell or progenitor cell identity. With various genetic mutants he was able to alter the ratio of symmetric and asymmetric cell division and was able to provide some support for the asymmetric DNA strand segregation. However he also presented the uncertainties of the system to prove that all stem cells behave according to the principles of this hypothesis. Self renewal and how this process is regulated by intrinsic and extrinsic factors is a central problem in stem cell biology. Self renewal is the key biological property that distinguishes a stem cell from any other cell type and it must be dysregulated in cancer.

Keith Humphries described studies of the role of Hox genes in regulating self renewal and how this process can be harnessed for stem cell-based therapeutics as well as how deregulation occurs in leukemia. Forced expression of Hox B4 results in enormous expansion of HSC number in vitro and in vivo, however enforced expression in progenitor cells cannot convert them into stem cells. He also made a point of distinguishing between self renewal potential and the actual execution of the self renewal program. He also described how Hox genes altered because of chromosomal fusion to partner genes can co-operate to disrupt self renewal and block differentiation resulting in leukemia.

Self renewal was also the theme of a presentation by Michael Clarke. He described studies on the gene Bmi1 which is implicated in the self renewal of several stem cells including HSC because KO mice rapidly lose HSC after birth and succumb to bone marrow failure. Bmi1 represses a number of genes including p53, p16, and p19. When each of these KOs is combined with Bmi1 KO, they are able to provide partial rescue of the HSC defect. Interestingly when all three downstream genes are mutated, the BM failure is rescued and HSC number restored. Careful analysis of these animals indicates that early progenitor cells gain self renewal potential, a property such cells have normally lost. If these animals are followed over time or are serially passaged, leukemia often arises. These studies begin to address the question of whether leukemia arises from the stem cell compartment or from the progenitor pool whereby these cells acquire self renewal potential.

The Forbeck Scholar Ben Ebert described an interesting series of studies to understand the basis for 5q- MDS. This disease is characterized by abnormal differentiation. He focused on the minimally deleted region and carried out a shRNA screen to determine if any of the genes within the deleted interval affects differentiation. He focused on erythroid differentiation and identified one gene RPS14 that blocks erythroid differentiation. He discussed other ribosomal genes that are also involved in other hematological diseases suggesting a link between this family of genes and erythroid maturation. He also described studies where he developed a genetic signature of erythroid gene expression that was predictive of MDS patients who respond to specific drug treatment. Ultimately such studies have enormous value in developing more effective therapeutics.

Irv Weissman described a wide range of studies on HSC biology from early development, to the effects of aging on HSC and leukemogenic alterations in HSC. He carried out a lineage fate mapping study using genetically marked chimeras. He showed that the yolk sac blood island is not clonal. Interestingly he also showed that the endoderm derives from a limited number of precursors. For example the intestine seems to derive from 12-14 precursors in large clonal patches. He also described the intrinsic and extrinsic changes that occur in HSC as they age. Aged HSC produce fewer lymphoid cells and more myeloid cells. As well the self renewal potential of the myeloid cells is increased. Collectively this insight into aged HSC shows that the age-related alterations in immune function can be traced back to HSC rather than age effects on the mature immune cells themselves. The alterations in HSC cycling, self renewal and the increased propensity for myeloid development including precursors that retain self renewal capacity provides mechanistic framework that explains the increased incidence of AML with age. Finally, he discussed the pathway of leukemogenesis from the normal cell of origin to the appearance of fully leukemic blasts. He argued that the initiating events must occur in HSC but that these are “preleukemic” in that they differentiate normally; the clone sustains additional genetic alterations. The mutations that occur leading to the generation of an LSC might occur in myeloid progenitors that now possess self renewal, thereby acting as self renewing leukemic stem cells. He proposed that examination of the hematopoietic system, especially the stem cell compartment by sequencing might uncover the number and sequence of pre-leukemic genetic alterations that occur before AML arises.

Amy Wagers presented a series of studies aimed at understanding the biology of murine HSC. She described a series of parabiosis experiments to examine HSC migration in a non-transplant situation and showed that while many HSC cycle every day, it still takes 7 to 16 weeks to obtain full donor chimerism. She then reported on studies to identify HSC-specific genes involved in mobilization and discussed her work on EGF1 which is differentially expressed to high level in long term repopulating HSC compared to multipotential progenitors. Knockout of this gene results in enhanced mobilization and HSC cycling. Thus EGR1 acts to limit HSC cell cycle when mobilized linking HSC proliferation and mobilization. She argued that regulators of this type are needed to ensure that HSC that are mobilized from their niches do not cycle since HSC that is the activity of HSC on regulating the osteoblasts that are involved in niche creation. The HSC that cycle do not home back to their niches. She also described her studies on the role of specific classes of osteoblasts in the niche (especially osteopontin expressing) and how they respond to.

Forbeck Scholar Carla Bender-Kim described her studies on the identification of a putative lung stem cell. These cells were identified on the basis of cell surface marker expression using an assay of napthalene challenge. Most lung cells are damaged with this agent, but the stem cells are spared and able to contribute to lung regeneration. She termed these cells the bronco-alveolar stem cell (BASC). She then targeted expression of the k-ras oncogene to the BASC and showed that their number increased dramatically before tumor appearance. She showed that ras expression could only generate tumors if it was expressed in the BASC cell types. She showed that the tumor stem cells within the generated tumors could self renewal as assayed by serial passage. Collectively her studies point the way to characterize in more detail the BASC and how they play a role in tumorigenesis. Finally they provide the basis to identify the human equivalent cell types in human lung cancers. The majority of our understanding of cancer biology has come from studies in murine models of cancer. However, there is need to understand neoplastic processes in primary human cells because of subtle differences between the two species.

John Dick discussed studies examining the hematopoietic stem cell hierarchy of primary human normal and leukemic cells using xenotransplantation systems. He showed that there were significant similarities between HSC and LSC, both contained stem cells with variable capacity for repopulation and self renewal organized as a hierarchy. The heterogeneity in LSC repopulation capacity poses a challenge to discover the molecular basis for this variable capacity and for the development of therapies that target the most quiescent LSC. He discussed one approach to eradicate LSC by targeting their ability to traffic in vivo and inducing them to differentiate. His group treated AML with specific monoclonal antibodies to CD44 and showed that they could interfere with LSC migration to the bone marrow and alter their stem cell fate via induction of differentiation. He also described a series of experiments to create an experimental system of human leukemogenesis by transducing normal human stem/progenitor cells with viral vectors expressing oncogenes including MLL fusion proteins. Upon transplantation into the xenograft models, all mice developed human ALL and AML. Such models could be powerful tools to study the leukemogenic process in primary human cells. Finally he discussed recent studies to determine the identity of the CSC that underlies human colon cancer. He showed a robust xenograft assay for the colon cancer initiating cell (CC-IC) that could be used to quantify their number. In addition the CC-IC could be enriched by sorting on the basis of CD133 expression. The normal cell within which brain cancer arises is not well understood.

Luis Parada described a series of experiments in murine models where he targeted oncogenes to be expressed in specific neural cell types. He provided evidence that the hippocampus retains cells with neurogenic potential, these cells express the marker GFAP. By targeting a GFAPcre marker to these cells, he showed that a stem cell defect occurred if cells expressing GFAP were ablated. He then went on to show that if the oncogene NF1 was expressed only in GFAP expressing cells in the context of a p53 deletion, that changes occurred in the sub ventricular zone before the appearance of tumor. He looked earlier in development at radial glial cells and showed that he only obtained brain tumors when the radial glia were targeted with NF1 expression as opposed to targeting other parenchymal cells. He then went on to examine the involvement of other regulatory genes to determine their role in oncogenesis as well as the sequence of activity. For example he showed that p53 mutations need to occur before the action of NF1 or no tumors developed, the addition of PTEN leads to alterations in neural stem cell function prior to frank tumor development.

Craig Jordan discussed his accumulating evidence for molecular distinctions between malignant stem cells in leukemia and normal hematopoietic stem cells. Prominent among these differences is activation of NF-kB and PI-3K, but other features such as increased oxidative state, immunoproteasome and gammaH2Ax suggest a cell under stress. One of the issues raised was whether the cancer stem cell has acquired the capacity to tolerate stress or has accumulated changes that themselves induce a state of cellular stress. Intervening to exploit that state is one hypothetical means of developing cancer stem cell therapies that may not similarly be toxic for normal stem cells. Craig showed his results with several agents that have a differential effect on leukemic stem cells compared with normal HSC. One of these, parthenolide, is moving forward toward clinical testing. The sequence of events and cell types that are involved in leukemic progression are poorly understood. This knowledge would aid in the more effective design of new therapeutics.

The Forbeck Scholar Catriona Jamieson presented data showing the sequence of events in the evolution of human CML from chronic phase to blast crisis. She showed that the phenotype of the LSC changes during disease progression so that in the blast crisis phase the LSC shows properties consistent with their evolution from a committed progenitor population. She then examined the molecular alterations between the CML LSC and the normal committed progenitors which they resembled and showed the inappropriate activation of the WNT pathway. This result opens the way for a new avenue of therapeutics that target the LSC. The most lethal component of cancer is when the tumor becomes metastatic, spreading from the primary site. A major question is whether every cell that migrates from a tumor has the potential to initiate metastatic growth.

Ann Chambers discussed a series of studies where she examined the proportion of cells that are able to initiate a micrometastasis and how many of these actually grow into a tumor at the distal site. She described a classical fluctuation analysis to calculate the rate of metastatic cell generation in a well studied murine model. She showed that only 2% of primary cells generated a micrometastasis and only 1% of those actually generated a tumor. She showed that majority of the micrometastatic cells remained dormant. Even in a highly metastatic cell line, a large proportion of the cells are not metastatic. She argued that the generation of metastatic cells is very rapid and not likely due to mutational events, rather due to epigenetic changes within the population of cells. Her presentation sparked significant discussion as to whether many of the micrometastases are inherently able to form tumors but cannot due to an inhospitable microenvironment, or whether many micrometastatic cells are not CSC and only the CSC within a tumor are able to initiate metastatic growth. The group felt that much effort should be made to identify and purify the metastatic cells within tumors.