In 1965, Sydney Brenner settled on Caenorhabditis elegans as a model organism to study animal development and behavior for reasons that are now well known (Brenner, 1973, 1988). This soil nematode offered great potential for genetic analysis, partly because of its rapid (3-day) life cycle, small size (1.5-mm-long adult), and ease of laboratory cultivation. One might imagine how the ability to grow thousands of animals on a single petri dish seeded with a lawn of Escherichia coli as the food source had a certain appeal to a bacteriophage geneticist such as Brenner. Indeed, the 300—350 progeny produced by a single animal is even greater than the burst of progeny produced by a T4 phage upon lysis of its E. coli host. The natural C. elegans mode of inbreeding by the self-fertilizing hermaphrodite combined with the ability to cross hermaphrodites with males (Fig. 1) offered conveniences previously enjoyed only in plant genetic systems such as Zea mays, in which crossing or selfing can be manipulated at will. Other key features were the nematode's small genome (only 20 times that of E. coli) and anatomical simplicity (>1000 cells), including the 302-cell hermaphrodite nervous system. With a nervous system that small, Brenner proposed that its complete circuitry could be determined by serial-section electron microscopy, a vision realized 20 years later (White et al. 1986, 1988). The ultimate goal was to determine the role of each gene involved in neural development and function.
An important reason C. elegans was chosen for study was that high-quality electron micrographs had been obtained from specimens of this species by Nichol Thomson, who was hired by Brenner in October, 1964. Initially, Brenner began reconstructing the nervous system by hand. He thought it might be possible to discern some principles of neural wiring by concentrating on one small part of the nervous system, so he started with the retrovesicular ganglion (RVG) at the anterior end of the ventral nerve cord. By the beginning of the 1970s, he began writing software for computer reconstruction of cell morphology from tracings of serial-section electron micrographs. Other investigators characterized responses to chemoattractants (Dusenbery 1973; Ward 1973) and the sensory ultra-structure of the wild type (Ward et al. 1975; Ware et al. 1975) and of mutants defective in chemotaxis (Lewis and Hodgkin 1977). Ultimately, the wild-type reconstructions showed all the connections of all the neurons in the hermaphrodite nervous system; the 381-cell male nervous system has been partially reconstructed at the EM level (White et al. 1988). Analysis of the wild-type circuitry allowed detailed models of how neurons function together to generate behavior. Furthermore, comparison of neural ultrastructure at different developmental stages revealed some surprising examples of developmental plasticity (Jorgensen and Rankin, this volume). In recent years, the wild-type circuit diagram has provided the foundation for interpreting the phenotypes of behavioral and locomotory mutants and the spatial deployment and function of neurotransmitters (Driscoll and Kaplan; Rand and Nonet; both this volume).
A criticism raised by early skeptics was that the animal has few morphological and behavioral traits. The organism was viewed as an almost featureless tube that moves forward or backward in a simple sine wave. It was suggested that with so few visible characters, it would be difficult to distinguish the functions of genes by mutant phenotypes, thus diminishing the power of genetics as a tool to reveal genes of interest. This criticism was somewhat dampened by the publication of the first C. elegans genetic map (Brenner 1974) containing more than 100 genetic loci dispersed over the six chromosomes, all of which are behavioral or morphological markers.
It was soon made clear that most of the interesting features are inside the animal. The transparency of the body, the constancy of cell number (eutely), and the constancy of cell position from individual to individual have perhaps been the most unique advantages offered by this organism for the study of development. It is a nearly ideal specimen for observation with differential interference contrast microscopy, and virtually every cell in the body is accessible to laser microsurgery. Eutely is of fundamental importance to the detection and reliable scoring of mutant phenotypes that alter cell lineages. The complete wild-type cell lineage from fertilized egg to adult was determined by observation of cell divisions and cell migrations in living animals (Sulston et al. 1988). This landmark provided a foundation on which much of the research in this book is based.
Brenner's overall plan could not be accomplished with the technologies available at the time. Major technical advances, such as the cloning and physical mapping of virtually the entire C. elegans genome, the development of transposon-tagging, reverse genetics, germ-line DNA transformation, genetic mosaics, and laser microsurgery, were essential to maintain and expand the usefulness of this model. Recently, Epstein and Shakes (1995) have compiled the most comprehensive collection of methods and information resources to date for analysis of C. elegans, including all of the methods mentioned above.
The early molecular genetic approach to C. elegans biology necessarily concentrated on genes encoding biochemically tractable products. This requirement effectively eliminated genes affecting neural development and function, but the genetics opened the door to address the problem of muscle assembly and function. The unc-54 gene, encoding the major body-wall myosin heavy chain, was chosen for study because its 210-kD product is one of the most abundant proteins in the body, and the protein was readily analyzed by SDS-PAGE (Epstein et al. 1974). A large number of mutant alleles with an easily recognized slow-paralyzed phenotype had been collected in Cambridge, including recessive, temperature-sensitive, and dominant alleles. One semidominant allele, e675, is a small deletion that results in a stable messenger RNA and a stable protein lacking 100 amino acids near the carboxyl terminus. The difference in mRNA size was crucial for identifying unc-54 complementary DNA clones, which were constructed from cDNA fragments synthesized from unc-54 mRNA that had been partially purified on sucrose gradients (MacLeod et al. 1981). Although ribosomal RNA repeats were cloned and characterized in the same year (Files and Hirsh 1981), unc-54 was the first genetically identified C. elegans gene to be cloned. It had a major impact on the muscle field because it provided the first myosin heavy-chain sequence. Up to that point, biochemists were proposing to sequence myosin using protein chemistry. unc-54 also provided a probe to clone homologous genes in other organisms. Studies on unc-54 quickly broadened to other muscle genes and muscle protein components, and the analysis of muscle assembly and function has been an important component in the exploitation of this experimental model (Moerman and Fire, this volume).
The long-term problem of having to curate stocks of the wild type and many mutant derivatives was solved in 1969 when John Sulston developed convenient methods for permanent storage of nematode stocks. Methods for storing viable stocks of nematodes frozen in liquid nitrogen were similar to those used for mammalian cell lines (Sulston and Hodgkin 1988). Stocks stored in liquid nitrogen for 25 years have retained their viability, as have stocks frozen at −80°C for the past 12 years. Laboratory strains maintained in growing cultures for long periods have been found to diverge with respect to fecundity and life span. Also, spontaneous activation of Tc1 transposition has been observed. This problem, by no means unique to C. elegans, has been minimized by the ability to return to frozen reference stocks at any time. Furthermore, the cost of faithfully maintaining large collections of mutants and the potential loss or mislabeling of strains that can occur with repeated transfer of growing cultures have been substantially reduced.
Brenner's initial emphasis clearly centered on the development and function of the nervous system, but a similar genetic approach to other aspects of developmental biology soon produced results, including the study of embryogenesis (Vanderslice and Hirsh 1976; von Ehrenstein et al. 1979), sex determination (Hodgkin and Brenner 1977), and larval development (Cassada and Russell 1975). Developmental genetics gained momentum as the postembryonic cell lineage was completed (Sulston and Horvitz 1977) and progressed faster than the neurobiology for a number of years, but a distinct shift back to neurobiology, including developmental neurobiology, is evident in this volume. Knowledge gained from the molecular genetics of development has provided new tools for analysis of the nervous system (Antebi et al.; Ruvkun; both this volume). This illustrates the value of “high-connectivity models,” such as C. elegans, yeast, Drosophila, the mouse, and mammalian cells in culture, in which many different aspects of their biology are intensively investigated (National Research Council Committee on Models for Biomedical Research 1985). In such a system, knowledge gained in one area of research ultimately “connects” with research in other areas. This connectivity both expands and reinforces understanding and speeds research progress. The more research areas that are investigated in a given model system, the greater the chance for connectivity. A read of this volume shows that this ideal has been realized.
Whether by chance or by design, basic biomedical research in the past 30 years has concentrated on a relatively small number of model systems (primarily prokaryotic cells, yeast, protozoans, C. elegans, Drosophila, Xenopus, Mus, primates, and mammalian cells in culture). Although these are quite different from each other, an astounding degree of connectivity between them has been revealed in the past decade. The emerging parallels between the development of the body plan in nematodes, flies, and mice (Ruvkun, this volume), and the fact that similar proteins are used for programmed cell death in both nematodes and humans (Hengartner, this volume), provide two examples.
C. elegans II attempts to provide a core of knowledge that builds upon the first book produced by the C. elegans community, The Nematode Caenorhabditis elegans, edited by W.B. Wood and the Community of C. elegans Researchers (1988). This 669-page book has been the “bible” of C. elegans biology for nearly a decade, and it will continue to be a fundamental resource for years to come. It synthesized the description of the organism, including its genetics, anatomy, cell lineage, development, reproduction, neural anatomy, and basic features of its genome. Here we profile knowledge that has accumulated within the past decade to apply this model system to solving current biological problems.
Cold Spring Harbor Laboratory Press, Cold Spring Harbor (NY)
Riddle DL, Blumenthal T, Meyer BJ, et al., editors. C. elegans II. 2nd edition. Cold Spring Harbor (NY): Cold Spring Harbor Laboratory Press; 1997. Section I, The Biological Model.