Embryogenesis in Caenorhabditis elegans is a remarkably reproducible process, mapped out to the resolution of every single cell division in the wild-type embryo (a total of only 558 cells at hatching). Despite the small number of cells, the newly hatched larvae contain all of the major types of differentiated tissues found in other animals (e.g., intestine, muscles, germline, epidermis, and a nervous system). Because of its relative simplicity, the C. elegans embryo provides a valuable opportunity to understand the process of animal development at the level of individual cells.
The C. elegans embryo utilizes several mechanisms to produce early blastomeres that are committed to distinct patterns of development. These mechanisms include position-dependent cell-cell interactions, as well as the asymmetric expression of maternally provided transcriptional activators and repressors. We are interested in the mechanisms that lead to the initial diversification of cell fates during the first several divisions of the embryo. One key protein involved in this process is the PIE-1 protein, a small Cys3His–zinc finger protein that functions in the specification of certain somatic cell fates and in maintaining the germline fate. After each round of cell division in the early embryo, the PIE-1 protein is localized to one and only one cell until the 28-cell stage of embryogenesis, when it is expressed in the two primordial germ cells. During early embryogenesis, PIE-1 appears to act as a repressor that protects the germline lineage from differentiating in response to transcriptional activators that are present in the germline. Our recent work suggests that PIE-1 not only prevents transcription but also blocks chromatin remodeling in the germline. We are exploring the mechanism of PIE-1 localization and function.
In addition to the various mechanisms that lead to the distinct fates of certain early blastomeres, several studies have indicated that cells throughout the embryo share a common mechanism for linking the axis of cell division to cell fate. For example, most of the cell divisions after the eight-cell stage of embryogenesis are oriented along the anterior/posterior (a/p) axis, and essentially all of these divisions result in a/p daughter cells with different fates. Using a combination of reverse and forward genetics, we identified several components of the signaling pathways that control cell fate and cell polarity in early C. elegans embryos.
Several genes involved in polarity signaling in the embryo encode highly conserved proteins related to Wnt-signaling components. Wnt genes comprise a large family, including numerous vertebrate and invertebrate members, and encode secreted proteins important in cell-cell interactions. In the early C. elegans embryo, Wnt signaling acts in parallel with a phosphotyrosine-signaling pathway mediated by SRC-1, a C. elegans homolog of tyrosine kinase pp60 Src. Wnt and SRC signaling converge to regulate WRM-1, a homolog of human β-catenin. Mutations that stabilize β-catenin are found in numerous forms of cancer in humans. Our ongoing studies involve a biochemical dissection of signaling, as well as genetic screening, with a focus on understanding how these and other signaling pathways function in concert to provide the positional information necessary for patterning the developing embryo.
RNA Interference and Related Pathways
During our studies on C. elegans embryogenesis we became intrigued by some strange properties of the gene-silencing method we were using. This method was analogous to "antisense," wherein the microinjection of RNA complementary to the messenger RNA is used to block the activity of specific genes. However, although antisense methodology generally causes, at most, a partial block of gene activity, this approach caused remarkably robust interference in C. elegans. For example, we were surprised to find that the silencing effect could spread from tissue to tissue in the animal, and even more surprisingly, could be transmitted for multiple generations in the germline as an epigenetic factor.
The remarkable potency and long-lasting effects of RNA interference (RNAi) prompted us to investigate the RNAi mechanism. In a collaborative study with Andrew Fire (then at Carnegie Institution of Washington; now at Stanford University), we showed that double-stranded RNA (dsRNA) in our antisense RNA preparations was responsible for inducing interference. This exciting discovery soon led to the application of RNAi via dsRNA in other organisms. To date, dsRNA-induced interference has been reported to one degree or another in a variety of animals and plants. In C. elegans and now several other organisms, dsRNA is so potent at causing interference that simply soaking animals in RNA, or feeding animals dsRNA, is sufficient to induce gene-specific interference.
In my laboratory we have identified several mutant strains that are resistant to RNAi. One class of mutants exhibits an activation of the endogenous transposons, selfish DNA elements, in the genome of the animal. This suggests that one in vivo function of the silencing mechanism is to silence the transposons. Molecular and genetic studies of RNAi, and of similar gene-silencing mechanisms in other organisms, have revealed underlying conservation of the silencing mechanism. For example, the rde-1 gene, required for RNAi in C. elegans, has homologs required for gene-silencing mechanisms in Drosophila, plants, and fungi. These and other findings indicate that an ancient mechanism underlies the RNAi process, and recent work has shown that related forms of gene regulation are important in development and disease in humans.
In C. elegans, genes related to rde-1 (members of the Argonaute gene family) are required for development, for fertility, and for proper chromosome segregation. Several studies have identified millions of endogenous small RNAs, and we have found that classes of these small RNAs interact with distinct members of the Argonaute protein family. For example, more than 15,000 small RNAs called 21U-RNAs, encoded in two large clusters on chromosome 4, comprise the piRNAs of C. elegans and engage the Argonaute protein PRG-1.
Another even larger class of small RNAs, termed 22G-RNAs, engage at least two functionally distinct Argonautes, CSR-1 and WAGO-1. CSR-1 engages 22G-RNAs targeting actively expressed mRNAs and is required for choromsome segregation during embyrogenesis. WAGO-1 and its associated 22G-RNAs target transposons, pseudogenes and other repetitive sequences, and appear to be required to silence these targets in the germline. Two other C. elegans Argonautes engage the microRNAs involved in developmental control of gene expression.
It appears likely that the expression of nearly the entire genome is monitored by these Argonaute-mediated small RNA pathways, but we are just beginning to unravel their complexities. Understanding these pathways will not only lead to better genetic tools for manipulating and studying gene function but will also shed light on ancient and highly conserved gene regulatory mechanisms. Ultimately, it may be possible to harness these natural mechanisms to modulate the expression of disease-related genes.