Monday, May 19, 2014

Complex gene action in a simple organogenesis model

ResearchBlogging.orgThe adult hermaphrodite of the roundworm, Caenorhabditis elegans, has only 959 cells, but even with these few cells, the animal has distinct tissues and organs, including muscles, skin, intestine, a nervous system, and an excretory system. The developmental origin of each of the 959 cells is largely invariant and is described by the cell lineage of the animal—the observed pattern of cell divisions from embryogenesis to adulthood. Further, the fate of each of these cells is also largely invariant, so we know for this animal where every cell comes from and what it will make in the adult. This information has made genetic analysis of development in C. elegans exceptionally fruitful.

The egg-laying system of the worm is comprised of two tissues, the somatic gonad that will form the uterus, and the vulva that will form the opening through which eggs are deposited and sperm can enter. Because the egg-laying system consists of cells from two different tissues, it can be considered a simple model for organogenesis, the process of organ development, which requires coordination of cell movements, divisions, and differentiation. Research on the signal from the somatic gonad that induces vulval cell divisions has been instrumental in illuminating the Ras cell signaling pathway, which is mutated to hyperactivity in a high percentage of human tumors.

In their studies of this simple organogenesis model, Rajakumar & Chamberlin identified multiple complex roles in multiple tissues for a single gene, egl-38, during development of the egg-laying system. We often see the reuse of a single gene, such as Sonic hedgehog, in multiple contexts during development, though this case is slightly different. Rather than distinct tissues and different processes, the roles of egl-38 involve different tissues, but in the process of forming the same organ. This developmental logic also appears to be conserved in other animals, as the related mouse Pax2 gene acts in reciprocal interactions between tissues during the branching organogenesis of the kidney.

In animals bearing the originally identified allele of egl-38, n578, cells in both the uterus and the vulva have defects, but are not completely abnormal. For example, the vulva cells fail to express LIN-3, an important signaling molecule, but do express other genes characteristically switched on in those cells. Nevertheless, the vulval cells fail to migrate properly and block the connection between the vulva and uterus. Likewise, uterine cells that should remain close to the vulva inappropriately migrate away.

One surprising result of the study is that different mutant alleles of egl-38 behave very differently from one another. Other alleles can cause significant defects in the hindgut and excretory system of the worm, but have little to no effect on the egg-laying system of the worm. It is not uncommon for alleles to differ in the severity of their effects—geneticists often can generate an “allelic series” based on these differing effects. For example, one tissue may be more sensitive than another to the partial loss of a gene’s function, so defects appear in that tissue fairly commonly, whereas another tissue may only display a defect when most or all of a gene’s function is disrupted. But in the case of egl-38, the allelic series is not a simple progression of defects appearing in more and more tissues as the strength of the allele increases. Instead, different alleles have qualitatively different effects, suggesting that the regions of the gene product affected by each mutation act in independent (or semi-independent) processes.

By far the most elegant experiment in the paper is the identification of which tissues egl-38 is required in to properly exert its function. Previous work had suggested that the LIN-3 signaling molecule, which is produced by the vulval cells, is important for proper migration of the uterine cells. Therefore, the egl-38(n578) defect in the uterus could conceivably be a secondary consequence of the loss of LIN-3 expression in the abnormal vulval cells. In a mosaic analysis, an animal is generated in which some cells carry the mutated allele of a gene and others carry a normal allele, and you observe how the two types of cells behave. In this paper, the two tissues are derived from distinct lineages of cells, and so generating an animal that lacks egl-38 gene function in one or the other is relatively simple [in C. elegans the mosaic is made using a mutant animal that carries an unstable extrachromosomal copy of the normal gene—cells that retain the copy are genotypically wild-type, whereas cells that lose the unstable copy are genotypically mutant; different techniques are used in other experimental systems, but the general principles of mosaic analysis are broadly applicable]. In animals with normal egl-38 function in both tissues, the vulva and uterus develop normally; in animals that have lost egl-38 function in both tissues, the vulva and uterus develop abnormally. And in the “money” experiment, animals that lose egl-38 function in just the vulva have an abnormal vulva but a normal uterus; and animals that lose egl-38 function in just the uterus have an abnormal uterus but a normal vulva. This result indicates that the function of egl-38 is cell autonomous—that is, the genotype of the cells in the tissue dictates the phenotype of the tissue. Far from being intuitively obvious or expected, this finding contradicts the model that the uterine defect is a secondary consequence of loss of LIN-3 in the vulva. Instead, egl-38 is required in the vulva for proper vulval development and in the uterus for proper uterine development. As beautiful as the mosaic experiment is, the function of egl-38 does appear to have at least one more layer of complexity. A final experiment shows that increasing signaling through the LIN-3 pathway can bypass the egl-38 uterine defect. That is, hyper-activating the signal that normally comes from the vulva and is dependent on egl-38 can compensate for the loss of egl-38 in the uterus, indicating that there appears to be both a vulval-dependent and vulval-independent role for egl-38 in the uterus.

As a final note, one reason I really like this paper and assign it to my students to read is that it is chock full of a diverse set of important techniques in developmental biology: reporter genes (like GFP), mosaic analysis, ordering genetic pathways by epistasis. After discussing these sorts of approaches in class, this paper demonstrates how effectively and elegantly they can be applied to dissect a developmental process.