Rolls, Stein, and colleagues (page 29) adapted a screening approach, which had previously been used only in yeast, to identify novel nuclear envelope proteins in mammalian cells. This technique could have a wide range of applications in cell biology and proteomics. The screen, which involves cloning a library of mammalian cDNA sequences into an expression vector with a green fluorescent protein (GFP) tag, allows the cloning of genes whose products have specific intracellular localization patterns. Cells are transformed with pools of cDNA-carrying plasmids, and sib selection allows enrichment of pools that show a desired pattern of fluorescence.

After demonstrating that known proteins localized correctly in the assay when fused to GFP, the team screened half of a cDNA library for novel nuclear envelope proteins. The screen uncovered a multiple membrane-spanning protein, subsequently named nurim, which is targeted to the nuclear rim but does not localize to nuclear pores or interact directly with lamins. The binding of nurim to the nuclear envelope is easily saturable, suggesting that it interacts with specific proteins rather than membrane lipids. “We are interested in identifying interaction partners of nurim. We hope that they will give us a clue as to why nurim is bound so tightly to the nuclear envelope, and point to its possible function,” says Tom Rapoport, senior author on the paper.

Though the new technique is not the lab's focus, others are likely to find it useful. Proteins with a variety of distribution patterns can be identified this way, and Rapoport adds that “there are many different variations on this kind of visual screen,” which would allow its application to large-scale functional genomics studies.

Starting on page 219, Le et al. describe the recycling of E-cadherin, which may be an important mechanism for regulating the activity of the receptor during cell–cell adhesion. Cadherins are important in development, tissue remodeling, and tumorigenesis. Though cadherin gene regulation has been studied, relatively little is known about the mechanisms that modulate cadherin adhesion.

To trace E-cadherin internalization, the researchers biotinylated cell surface proteins, then allowed endocytosis to proceed under various conditions before stripping exposed biotin with glutathione. Receptors that had been internalized were protected from stripping and remained biotinylated, allowing detection of the internalized receptor pool. Le et al. found that a population of E-cadherin is constantly recycled from the cell surface to an intracellular compartment but is not targeted for degradation. The size of the recycling pool is significantly larger in cells lacking stable cell–cell contacts than in cells growing in confluent monolayers, and the reformation of disrupted cell junctions is inhibited when recycling is blocked. “Trafficking, moving proteins in and out of the cell membrane, is emerging as a major mechanism for regulating the cell surface functions of channels, transporters, and receptors. It now seems that cell adhesion can also be regulated in this way,” says Jennifer Stow, senior author on the paper. E-cadherin recycling appears to take place through clathrin-coated pits, and the team is now studying cell lines with inducible defects in clathrin-mediated endocytosis and examining the recycling of other components of the cadherin complexes.

In parallel studies of two different model systems, researchers have discovered a new family of dynein light chain proteins that may modulate the activity of the molecular motor in a wide range of species. Dynein has a key role in mediating intracellular transport, but its ability to interact with a large number of other proteins has complicated efforts to identify regulatory components of the protein complex.

In the new work, reported on page 165, Bowman et al. used a genetic screen of Drosophila to identify mutants with posterior sluggishness, a phenotype associated with defects in axonal transport. One mutant, dubbed roadblock, exhibited progressive sluggishness, leading to complete posterior paralysis, and lacked imaginal tissue, suggesting a strong mitotic defect as well. Independently, a search for dynein components in Chlamydomonas was being done with a biochemical approach, purifying and partially sequencing the LC7 component of outer arm dynein. “We had mystery sequence for a [fly] gene that gave a phenotype while Steve King had a sequence of Chlamydomonas LC7. We became aware of each other's existence… when we noticed some sequence homologues in common,” explains Lawrence Goldstein, an author on the work. Disruption of roadblock demonstrated that the gene product has a direct role in dynein function, and additional homology searches uncovered a roadblock/LC7 family of dynein light chains in several species, including humans. The findings add to a growing body of data showing that light chain mutations in dynein tend to cause less severe and more diverse phenotypes than heavy chain mutations, suggesting that these components have a regulatory role.

Using electron microscopy and immunofluorescence, Brunet et al. (page 1) have demonstrated that, in contrast with mitosis, the formation of a bipolar spindle during the first meiotic division in mouse oocytes occurs in the absence of kinetochore fibers. In the first meiotic division, homologous chromosomes are segregated while sister chromatids remain together, suggesting that the molecular mechanisms of meiosis and mitosis differ considerably. Previous studies of spindle formation have generally focused on mitotic cells.

Electron microscopy revealed that during the first several hours of meiotic metaphase in mouse oocytes, microtubule ends interact with the chromosomes, but not with the kinetochores. Disrupting the prometaphase spindle with nocodazole for up to 3 h did not delay polar body extrusion, indicating that the continuous presence of this structure is not required. At the end of the first meiotic M phase, microtubules begin to interact with kinetochores, and the authors suggest that this interaction controls the timing of the first meiotic division. The observations contrast sharply with studies of mitosis, in which stable kinetochore–microtubule interactions are seen as soon as the nuclear envelope breaks down.

The formation of kinetochore fibers at the end of meiotic metaphase does not require new protein synthesis, but further details of the process remain unknown. “There is no clear candidate for the end attachment of the microtubule to the kinetochore. We have to look for proteins that could be involved in this interaction,” says Bernard Maro, senior author on the paper.

Building on earlier work, that identified DAP-kinase as a potential mediator of apoptosis signals, Cohen et al. (page 141) have now determined that the enzyme lies along both the TNF-α and Fas-mediated apoptotic pathways. The researchers also found that the DAP-kinase death domain is required to transduce the apoptotic signal. The protein was originally identified based on the ability of DAP-kinase antisense RNA to prevent cell death induced by IFN-γ, but its significance and functional position in the apoptotic signaling cascade remained unknown.

After determining that DAP-kinase antisense RNA inhibits cell death mediated by the distinct Fas and TNF-α pathways, the team demonstrated that the death domain of the enzyme is required for signaling, and can act as a dominant-negative inhibitor of apoptosis. Overexpression of mutant DAP-kinase causes cell death, an effect which can be blocked by bcl-2 and inhibitors of some caspases, but not by dominant-negative mutants of components of the TNF and Fas receptor complexes, indicating that the kinase acts downstream of the receptors and upstream of other caspases. The team is now attempting to isolate other components of the pathway using a two-hybrid screen.

Though it may be a central death effector gene, inhibiting DAP-kinase does not completely block apoptosis. Senior author Adi Kimchi explains that “our interpretation for the partial inhibition of apoptosis is that the signaling pathway has already begun to diverge at the point where DAP-kinase functions,” but adds that “the other branches of the pathway could differ completely in their nature and may not involve analogous proteins.”