Cédric Blanpain (organizer), John E. Dick, Tariq Enver, Elaine Fuchs (organizer), Magdalena Götz, John B. Gurdon, Brigid Hogan, D. Leanne Jones, Jürgen Knoblich, Thomas (Tom) Rando, Timm Schroeder, Benjamin (Ben) Simons, Austin Smith, Azim Surani, Shahragim Tajbakhsh, Elly Tanaka, Maarten van Lohuizen, Irving (Irv) Weissman, Shinya Yamanaka, Leonard (Len) Zon.
by Cedric Blanpain
20 – 25 October, 2014
Stem cells are essential for the development of the different tissues and organs, for sustaining the daily replacement of the cells that are continuously lost and for the repair of most tissues following injuries. Stem cells can also serve as the cancer cell of origin and cancer may contain cells with higher renewal properties called cancer stem cells. This seminar gathered stem cell biologists working on various model systems ranging from drosophila to human stem cells, studying the connection between stem cells and cancer and modeling the dynamics of stem cells. The Fondation des Treilles offered a unique setting for illuminating discussions across different fields of science, defining the key outstanding questions in stem cell biology and how stem cells could be used in the future to better understand and treat human diseases.
Stem cells; development; reprogramming; pluripotency; regeneration; cancer stem cells;
Austin Smith gave a presentation on the nature of embryonic stem cells, where he considered three issues. What is the identity of embryonic stem cells – are they an artificial cell state created in the laboratory or do they represent a precise stage of cellular development in the early embryo that can be “captured” in vitro? Is there an evolutionary force that can account for propagation of embryonic stem cells since self-renewal is not normally observed in the early embryo? Can authentic embryonic stem cells be derived from other mammals than rodents, in particular from human?
The general theme of John Gurdon’s work is to understand the resistance of nuclear transfer to the reprogramming activities of eggs and oocytes of Amphibia. John Gurdon spoke in particular about his recent work that demonstrates a brief window of opportunity for transplanted somatic cell nuclei, during the cell cycle, to escape from the stabilizing effects of cell differentiation and to undergo nuclear reprogramming. A special window of opportunity exists briefly during mitosis when chromatin becomes momentarily released and a new opportunity for reprogramming nuclear function appears at that time. The mechanism seems to involve a loosening of chromatin and a major exchange of transcription factors. This “mitotic advantage” may be a general characteristic of normal early development.
Shinya Yamanaka discussed two issues about iPS cells. First, he demonstrated how aberrant DNA methylation during iPS generation affected differentiation propensity. Second, he shared the results of genetic and epigenetic analyses of retinal pigmental epithelial cells derived from human iPS cells, which had been transplanted into a patient suffering from age-related macular degeneration as the first-in-human clinical trial using iPS cell technology.
Jurgen Knoblich presented his work on brain tumor formation in the fruit fly Drosophila. Defects in stem cell lineages can lead to the formation of hyper plastic tissue overgrowth that resembles human tumours in a remarkably precise manner. The group has used this model system to identify the rate-limiting and irreversible steps of tumor formation and found specific epigenetic mechanisms that are most critical for tumorigenesis.
Marteen van Lohuizen showed how Polycomb repression sets thresholds for differentiation cues: loss of PRC2 in the intestine induces aberrant differentiation at the expense of stem cells. He further described testing the potential use of EZH2 inhibition in Glioblastoma using conditional shRNAi models. While initial beneficial response was observed, interestingly prolonged EZH2 inhibition caused the tumor cell to adopt in a epigenetic ‘escape’ aggressive tumor state rendering them insensitive to standard of care chemotherapy due to up regulation of pluripotency and DNA damage response networks. This indicates highly context dependent effects of EZH2, acting either to promote or suppress tumor progression.
Ely Tanaka spoke about limb regeneration as a model for understanding how complex, patterned tissues can be formed in the adult context. She showed that patterned limb regeneration involves the cooperation of connective tissue cells deriving from anterior and posterior parts of the limb that show different determination states. The anterior tissue has the competence to express FGF8 after limb injury, can only sustain the expression upon stimulation of the SHH pathway, but is in competent to express SHH itself. Complementarily, the posteriorly derived blastema cells are competent to expression SHH but not FGF8. This cross-stimulation of these two cell populations with different determination states is required for limb outgrowth as well as anterior/posterior patterning. Interestingly, global application of these factors without temporal regulation yielded appropriately differentiated skeletal elements, suggesting that other factors determine when regeneration should stop.
The zebrafish is an excellent model for studying cancer. Len Zon’s lab has created a transgenic zebrafish that marks the migrating neural crest in embryos with green fluorescent protein. This transgene is expressed until three days of development, and then is silent even in adults. When melanoma arises, which is neural crest derived, the tumor becomes green. With this transgene, the cancer can be visualized when it is one transformed cell. This cell expresses a neural crest program. Overexpressing the master regulator of neural crest, sox10, increases the rate of fluorescent cells and cancer. The promoter used has sox10 binding sites. An analysis of histone marks in melanoma shows that superenhancers are activated in the fluorescent patches that precede tumor formation. The zebrafish has therefore provided a method to visualize cancer at the one cell stage, and has begun to define early events in cancer development.
John Dick showed that gene signatures specific to either AML LSC or normal HSC were identified and found to be highly similar, defining a common stemness program. These stem cell signatures were significant independent predictors of patient survival in 4 large clinical databases of >1000 samples. Thus, determinants of stemness influence clinical outcome of AML across a spectrum of mutations indicating that many genetic abnormalities coalesce around stem cell properties. Secondly, studies were presented that found that pre-leukemic HSC and progenitors could be isolated from the blood of AML patients at diagnosis. Through genetic and functional approaches, it was possible to trace back the origin of AML to the initiating oncogene arising in HSC, leading to a clonally expanded pool of preL-HSC from which AML evolves. Third, a mechanism was uncovered whereby the duration of exit from quiescence is differentially regulated within human HSC subsets by pre-existent levels of CDK6 within quiescent cells. Experimentally accelerating the duration of quiescence exit in LT-HSC confers a competitive advantage without loss of function. Thus pre-leukemic clonal expansion could arise by simple perturbations of cell cycle kinetic control without affecting self-renewal processes.
Irv Weissman purified mouse and human hematopoietic stem cells [HSC] and modified the Lapidot-Dick us of antibodies to show that human AML leukemia stem cells[LSC] are mainly at the multipotent progenitor [MPP] stage of development. Exome sequencing of human AML to find patient-specific somatic mutations led he and Majeti to show within 21 patients with AML that the clonal progression of preleukemia occurs always in an HSC clone that accumulates the mutations, and the final, proliferative/self-renewing step such as K-ras or N-ras or Flt3 internal tandem repeat occurs at the MPP stage of these clones. The early mutations tend to be in epigenome modifying genes such as Tet2, Dnmt3a, CTCF, and Idh1/2. Comparing the gene expression of human HSC and MPP vs human LSC led to the discovery that all AMLs in humans overexpress CD47, a ‘don’t eat me’ signal for macrophages. This led to the finding that overexpression of CD47 occurs in all cancers, and is, to date, the single gene expression change common to all cancers. Weissman finished by showing that a humanized anti-CD47 produced by him and Majeti shrinks or eliminates all primary human cancers and their metastases transplanted into immune deficient mice. That antibody is currently in phase 1 clinical trials.
Timm Schroeder highlighted the importance of continuous long-term single cell quantification as a prerequisite for a mechanistic understanding of molecular stem cell fate control. He introduced his bioimaging and electronic image analysis approaches allowing the required generation of quantitative data on cell fate choices and molecular dynamics over many cell generations. Through these approaches, the expression of transcription factors can be quantified live and long-term in single stem cells, revealing surprising expression dynamics not compatible with long-standing text book models of how transcription factor networks control stem cell fates.
Tariq Enver presented evidence on the extent to which extrinsic and intrinsic transcriptional noise may contribute to its generation.
He showed that cells transiting lineage commitment boundaries show heterogeneity in expression of key lineage affiliated regulators presumably reflecting intrinsic noise in gene expression. These results suggest that individual cells may undergo cell specification using different gene expression trajectories which later converge as cells mature. In respect of extrinsic noise he showed that individual multi potent cells differ in overall transcription rate with fast transcribing cells displaying increased potency. Transcriptional speed correlates with mitochondrial membrane potential and thus stochastic segregation of mitochondria at cell division may contribute to functional cell heterogeneity.
Elaine Fuchs discussed her lab’s work on how extrinsic niche cues, particularly those eliciting changes in BMP, WNT and SHH signaling, triggers a cascade of chromatin-associated and transcriptional changes within skin stem cells that governs their activation during tissue development, homeostasis and hair regeneration. They’ve applied this knowledge in exploring how stem cells change as they exit their niche and embark upon a specific lineage program or alternatively participate in wound-repair following injury.
Their findings provide new insights into the normal process of stem cell activation during homeostasis and wound-repair, and in so doing, they began to realize that malignant progression hijacks these basic mechanisms which are essential for all tissues. They’ve now focused on how stem cell behavior goes awry during tumor progression. They’ve purified and characterized functional skin tumor-initiating cells at near homogeneity. They’ve also developed a new method to knockdown genes specifically in skin and oral progenitors, enabling them to screen not only the differences between these cancerous and normal stem cells, but also the myriad of gene alterations surfacing from the Human Cancer Sequencing project. They’ve also carried out whole genome-wide RNAi screens for oncogenic growth of stratified epithelia. Their screens have illuminated new oncogenes and tumor suppressors for squamous cell carcinomas, among the most prevalent and life-threating cancers world-wide that include cancers of lung, esophagus, breast, cervix, prostate, throat and oral tissues. Their findings are unearthing new targets for cancer therapeutics, as well as for delving deeper into understanding the mechanisms that underlie malignancy and metastasis.
Brigid Hogan addressed the topic of epithelial stem cells in the lung. Unlike many other organ systems, cell turnover is normally very slow in the lung yet there is an impressive capacity for repair after damage caused by toxic agents or infection. Evidence suggests that different regions contain different populations of stem and progenitor cells. In the proximal airways Krt5+ p63+ basal cells are the major source of reparative cells. There is growing evidence that this population is heterogeneous and changes in composition along the dorsal-ventral axis and with age. A three-dimensional organoid culture system has been developed to screen for factors driving the self-renewal and differentiation of the basal cells. A comparable organoid system has been used to identify mechanisms regulating the behavior of type 2 epithelial cells, the stem cells of the distal alveolar region. Here, evidence suggests that Pdgfra+ lipofibroblasts constitute a critical component of the stem cell niche and signal to the stem cells to regulate their differentiation into type 1 cells.
Beginning with a conceptual overview on the emergence of collective phenomena in the non-equilibrium dynamics of statistical ensembles, Ben Simons gave a presentation on how ideas from physics could be applied to study stem and progenitor cell dynamics in adult tissues. Using these approaches, our analysis of long-term clonal evolution and short-term in vivo live-imaging of mouse germ line revealed a process of stochastic stem cell loss and replacement. We discussed how these findings provide new insight into dynamical stem cell heterogeneity and the maintenance of other tissue types, including intestinal crypt.
Magdalena Götz presented the outcome of the first live in vivo imaging of adult neural stem cells revealing not only their behaviour in the normal healthy zebrafish brain, but also unravelled several mechanisms incasing the neuronal output after brain injury including changes in the mode of cell division. Towards installing such neuronal repair also in the injured mammalian brain novel mechanisms in direct reprogramming of glial cells into neurons aiding their metabolic conversion reaching more than 80% efficiency in vivo were presented.
Specification of primordial germ cells (PGCs), the precursors of sperm and eggs, is intimately linked with the initiation of unique germline reprogramming, necessary for resetting of the epigenome for totipotency and development. Azim Surani showed that PGC specification in mice requires three transcription factors; BLIMP1, PRDM14 and AP2g, which are necessary and sufficient for the germ cell fate. Germline reprogramming leads to an epigenetic ground state with the comprehensive erasure of DNA demethylation, which leads to the upregulation of transposable elements (TEs). Host defence mechanisms have evolved for their suppression. PRMT5, an arginine methyltransferase was recently found plays a role in the repression of TEs before piRNA biosynthesis, as well as during preimplantation development. Some of the TEs have been co-opted during evolution to regulate mammalian development and pluripotency network. Less is known about the human germ cell lineage. New approaches are being developed to study the mechanism of human PGC specification from pluripotent cells.
In the process of studying stem cell quiescence, Thomas Rando found that there is an important “activation” step when quiescence stem cells much first grow in size before undergoing their first division, and that process of activation results in a much greater time for quiescent stem cells to divide than for any subsequent division. We observed that in response to injury to a muscle, not only to stem cells in that muscle activate in order to begin the process of tissue repair, but muscle stem cells throughout the body respond by enlarging and displaying activation of the mTOR signaling pathway, into what we have termed an “alert” state, or G(Alert) as another state of quiescence. From G(Alert), muscle stem cells repair injuries for rapidly and effectively than from G0. We have found that this alert state is dependent on an active Hepatocyte Growth Factor (HGF)/c-met signaling pathway in the stem cells, and also that not only muscle stem cells but also mesenchymal, hematopoietic, and skin stem cells appear to enter an alert state in response to distant injuries and also show functional enhancement. Our current studies focus on identifying the systemic factors or signals that are conveyed from the site of injury to stem cell compartments throughout the body. We hypothesize that this process that reflects a kind of cellular memory and that is an adaptive mechanism to enhance tissue repair in the setting of recurrent injuries.
It remains unclear whether the cancer cell of origin controls tumor heterogeneity. Cédric Blanpain showed that conditional expression of oncogenic KRas together with p53 deletion in different epidermal lineages lead to different tumor heterogeneity depending on the identity of the cells initially targeted. Fractionation and transplantation of distinct population of tumor cells identified a population of cancer cells with enhanced self-renewal and multilineage differentiation properties. Transcriptional profiling allowed the identification of key regulators controlling tumor stemness and the maintenance of multilineage differentiation in skin tumors.
Leanne Jones spoke about “Age-related changes to stem cells and the stem cell niche, using Drosophila germline and intestinal stem cells as model systems. The central focus of the talk was new data indicating that ageing results in loss of tight junction (TJ) proteins from the Drosophila intestinal epithelium, concomitant with other aging phenotypes including increased ISC proliferation and a block in terminal differentiation. Targeted depletion of TJ proteins between differentiated cells leads to similar aging phenotypes in young animals, indicating that TJ integrity is an important non-autonomous regulator of stem cell behavior in this tissue.
Shahragim Tajbakhsh discussed studies on how the skeletal muscle cell lineage is established in the embryo and how Notch plays a permissive role in this process. He also presented data on single cell analysis of adult skeletal muscle stem cell divisions using micro-patterns where modulation of the topology of adhesion cues can dictate the choice between symmetric and asymmetric segregation of DNA and transcription factors.