Stem cells generate tissues during the development of multicellular creatures and maintain those tissues in adults. Understanding stem cell regulation is not only fascinating in its own right, but also is crucial for regenerative medicine. My lab studies mechanisms of stem cell self-renewal, cell fate specification, and reprogramming in the small nematode Caenorhabditis elegans. We have identified key regulators of stem cells and differentiation that are conserved in vertebrates, including humans; identified conserved pathways and molecular mechanisms driving fundamental biological regulation more generally; and developed a new model for analysis of reprogramming cell fates in vivo.
A Stem Cell Niche, Notch Signaling,and Regulation of a Stem Cell Pool
The concept of a stem cell niche was proposed in 1978 as an "external microenvironment" required for stem cell maintenance. In 1981, we discovered that the mesenchymal distal tip cell (DTC) provides the niche for germline stem cells (GSCs) in C. elegans (Figure 1). A major goal of my laboratory has been to understand at a molecular level how the DTC niche regulates stem cells—including the niche signal and the stem cell response to self-renew or differentiate. Our early work revealed that the DTC niche uses Notch signaling to drive GSC self-renewal and that GSCs reside within a cluster, or "pool," of stem cells. The Notch signaling pathway is broadly conserved among animals; its C. elegans version for GSC regulation is called GLP-1/Notch. Other researchers have found that vertebrate stem cells also rely on mesenchymal niches and Notch signaling for their maintenance. Moreover, the idea of a "stem cell pool" has gained traction in vertebrate tissues, including humans.
An RNA Regulatory Network Controls Self-Renewal and Differentiation
Over the past decade, we have elucidated key regulators that act downstream of Notch signaling within the stem cells and their progeny to drive self-renewal or differentiation, and we have discovered biochemical links among those regulators to reveal a molecular network of control (for simplified network, see Figure 2A). Central to this network are an "FBF" hub that drives stem cell self-renewal and two "GLD" hubs that drive meiotic entry in stem cell daughters. One of the most remarkable features of this network is that its key hubs regulate mRNA stability or translation. FBF, GLD-1, and GLD-3 are all conserved RNA-binding proteins, and GLD-2 is a conserved poly(A) polymerase. Another striking feature is that no individual component of the network is essential for self-renewal or differentiation; instead, removal of most single components simply shifts the balance between self-renewal and differentiation. That redundancy makes the network exceptionally robust and plastic.
FBF, A Conserved RNA-Binding Protein and Stem Cell Regulator
FBF (fem-3 binding factor) is a founding member of the conserved PUF (Pumilio and FBF) family of sequence-specific RNA-binding proteins. To maintain stem cells, FBF represses a battery of >1,000 mRNAs, many of which encode proteins regulating differentiation (Figure 2B). About 15 such mRNAs were discovered as candidates, and the rest were found genomically. One class of FBF target mRNAs encodes components of the meiotic program, including both master regulators of meiotic entry and meiotic cell cycle machinery. Another class encodes regulators of germ cell differentiation as either sperm or egg, and yet another class encodes more general differentiation regulators (e.g., mpk-1, the nematode homolog of ERK/MAP kinase). Most FBF analyses have been done in collaboration with Marvin Wickens (University of Wisconsin–Madison).
The FBF role in stem cell maintenance is broadly conserved, as are many of its target mRNAs. PUF proteins maintain GSCs in fruit flies, and they maintain neoblasts in planaria. PUF proteins are also present and functional in vertebrate stem cells, including human embryonic stem cells, but their role in vertebrate stem cell maintenance has not yet been tested. A comparison of FBF target mRNAs to targets of its human counterparts, which are called PUM, reveals hundreds of shared mRNAs, including those in the core nematode network (Figure 2A and 2B). The mRNA encoding ERK MAP kinase was validated as a conserved PUF target in nematodes and human embryonic stem cells (collaboration with Dr. Jamie Thomson [University of Wisconsin–Madison]). The target mRNAs shared by nematode FBF and human PUM encode key regulators of cell death (e.g., CED-4), aging (e.g., DAF-16/FOXO), and numerous canonical signaling pathways (e.g., Ras/MAPK, NFκB, Notch, and PI3/Akt) (Figure 2B). Therefore, not only are PUF proteins conserved, but their regulatory relationships may also be ancient.
PUF Proteins Can Repress mRNAs Via Two Conserved Molecular Mechanisms
PUF proteins bind specific regulatory elements in the 3′ untranslated region of their target mRNAs (Figure 3A). Their binding then leads to mRNA repression by either of two molecular mechanisms. One mechanism involves recruitment of a deadenylase complex to remove poly(A) and hence to destabilize or translationally silence the mRNA (Figure 3B). This mechanism was discovered first for yeast and human PUF proteins and appears broadly conserved. The second mechanism we first discovered first using FBF and nematode proteins, and then showed its conservation with human proteins. Briefly, PUF proteins (C. elegans FBF and human PUM2) form a ternary complex with an Argonaute protein (C. elegans CSR-1 and human Ago1-3) and a core component of the translational machinery, elongation factor eEF1A. Argonautes are best known for binding small RNAs and their role in microRNA-mediated mRNA repression. The PUF-Ago-eEF1A complex appears to repress mRNAs by stalling ribosomes within the open reading frame at a site roughly 100–140 nucleotides from the AUG start site of translation (Figure 3C). We speculate that PUF-Ago-eEF1A may be the first of many multisubunit regulatory complexes that include core components of the translation machinery.
The major idea emerging from our FBF work is that PUF proteins maintain stem cells as broad-spectrum mRNA repressors of differentiation. A challenge for the future is to learn whether the distinct PUF molecular mechanisms play specific roles within stem cells. An attractive model is that PUF repression uses these mechanisms in many organisms, perhaps in many tissues, to complement use of transcriptional regulators for stem cell and progenitor cell control.
Molecular Regulation of the Stem Cell Niche and Insights into the C. elegans Wnt Pathway
Understanding how stem cell niches are themselves regulated will enhance our ability to manipulate stem cells for regenerative medicine and to treat human disease. Yet, molecular controls of niche specification and maintenance are largely unexplored. We tackled this fundamental problem in C. elegans, where the DTC niche provides an unusually simple and well-defined niche. First, we learned that the Wnt pathway activates ceh-22 gene, which encodes the C. elegans Nkx2.5 homeodomain transcription factor, to specify the niche fate. An intriguing but speculative parallel is the role of Nkx2.5 in development of the vasculature, which serves as a niche for neural and hematopoietic cells in vertebrates. Second, we discovered a functional β-catenin not recognizable by its amino acid sequence, a finding that expands the definition of β-catenins and suggests that other divergent β-catenins await discovery. Third, our work and that of others revealed that the major C. elegans Wnt pathway is branched—one branch up-regulates β-catenin and the other down-regulates T cell factor (TCF). We also discovered that this branched pathway controls the ratio of β-catenin to TCF and that it is this ratio that drives Wnt-dependent transcriptional activation. If vertebrates similarly rely on the ratio of β-catenin to TCF, the pharmaceutical lowering of TCF might activate Wnt signaling, a prediction with important clinical implications.
The Sperm/Egg Decision and Its Chemical Reprogramming
We have begun to understand, for the first time in any organism, the mechanism that acts within germ cells to control the sperm/egg fate decision. One unexpected outcome of knowing these molecular regulators has been the discovery that adult germ cells are sexually labile—manipulation of its regulators can sexually transform the adult germline from spermatogenesis to oogenesis, or vice versa. Therefore, these regulators must remain active in adults to continue specifying germ cells as sperm or oocyte during the continuous production of gametes.
To provide a potentially general means to understand cell fate reprogramming, we have discovered a method to use small molecules for reprogramming germ cell fates. We add chemical inhibitors of the ERK/MAP kinase pathway to a spermatogenic animal and sexually transform its germline to make oocytes that can support embryogenesis and produce viable offspring (Figure 4). This reprogramming of germ cell fates in C. elegans occurs in stem cell progeny soon after they leave the niche and before overt differentiation. Reprogramming therefore does not occur by direct conversion of a differentiated spermatocyte into a differentiated oocyte, but instead takes place as stem cell progeny begin to acquire their differentiated fate outside the niche. To our knowledge, this is the first time that a cell fate reprogramming event has been mapped in vivo. These studies in nematodes provide a powerful model for analysis of molecular mechanisms that drive cell fate reprogramming in vivo. Our findings represent a paradigm that may facilitate pharmacological approaches to therapeutic cellular reprogramming in other organisms, including humans.
Mitosis/Meiosis and Sperm/Egg Choices Are Separate Regulatory Events
A fundamental question has been whether germ cells commit to the meiotic cell cycle (mitosis-meiosis decision) and to sperm or oocyte differentiation (sperm-oocyte decision) via one or two cell fate choices. If a single choice is involved, a male-specific or female-specific meiotic entry would lead necessarily toward spermatogenesis or oogenesis, respectively. If two choices are involved, it should be possible to experimentally separate from specification as sperm or oocyte. We investigated the relationship between these two choices with tools uniquely available in C. elegans. Specifically, we used a temperature-sensitive Notch allele to drive germline stem cells into the meiotic cell cycle and subsequently used chemical inhibition of the Ras/ERK pathway to reprogram the sperm-oocyte decision. Germ cells that had already entered meiotic prophase were nonetheless capable of being sexually transformed from a spermatogenic to an oogenic fate. This finding cleanly uncouples the mitosis-meiosis decision from the sperm-oocyte decision and suggests that this fundamental separation will hold true for germ cells throughout the animal kingdom.
A grant from the National Institutes of Health provided support for some of these studies on germline controls.
As of March 27, 2013