Cellular suicide pact

Dendritic spines cut loose to mature D eveloping dendritic spines need to hold on tight to the extracellular matrix as they seek out partnering neurons. Once they fi nd partners, however, the spine tips must sever their matrix ties to reshape into post-synaptic receiving stations, according to Tian et al. (page 687). Dendritic spine growth and maturation coincides with the expression of a neuron-specifi c transmembrane adhesion molecule called ICAM-5. Spines in mice lacking ICAM-5 grow more slowly but mature more quickly, suggesting that the adhesion molecule is needed for spines to fi nd partners but inhibits their subsequent maturation. Tian et al. now show that spine maturation is possible in neurons because much of ICAM-5 is cleaved into a nonadhesive form during synaptogenesis. This cleavage was promoted by neuron stimulation through receptors for the NMDA and AMPA neurotransmitters. Receptor activation caused ICAM-5 to be released from the actin cytoskeleton, which in turn promoted its cleavage. Release from the actin network might make it susceptible to destructive enzymes such as the MMP2 and MMP9 matrix metalloproteases, which were needed for ICAM-5 cleavage. Cellular suicide pact M ass suicides are not the reserve of cults, it appears. On page 567, Link and colleagues report that cells do it too. Cellular suicide—known as apoptosis—is necessary during development for sculpting the shape of organs and limbs. Most of the time, individual cell suicides are dotted throughout the tissue, and cell corpses are then cleared away by phagocytosis. Link et al., however, have now captured by live imaging a mass cellular suicide in fl ies. Epithelial wing cells died in a wave that swept across the developing wing in a matter of minutes. The dead cells were not phagocytosed but were instead swept into the wing veins, which drain into the body. The authors hypothesize that a released signaling factor or mechanosensory response propagates the swift en masse suicides. A mutant screen for defects in this collective cell death has not yet identifi ed any such instigators but has revealed novel cell death genes, including homeodomain-interacting protein kinase. This gene’s product functions in collective cell death and in more conventional examples of programmed cell death. The authors speculate that communal cell death might be a more widespread phenomenom. Massive cellular suicides might occur, for example, in mammals during development or after rapid hormonal changes, such as mammary gland involution after nursing and the shedding of the uterine lining at menstruation. SCAPER speeds and slows the cell cycle C yclin A promotes the progression of both S phase and mitosis. A new cyclin A binding protein, identifi ed by Tsang et al. (page 621), helps the S phase push but hinders mitotic progression. The new partner seems to favor S phase by holding cyclin A in the cytoplasm. To determine how cyclin A pushes forward different stages of the cell cycle, Dynlacht’s team looked for new binding partners. They fi shed out one especially strong candidate, which was associated with the ER, and named it SCAPER (S phase cyclin A–associated protein of the endoplasmic reticulum). Overexpression of SCAPER delayed progression of cells through M phase, suggesting that the protein blocks this cyclin A activity. But SCAPER deletion prevented cells from entering S phase in a timely manner, conversely indicating that it normally helps cyclin A in this case. This S phase delay, the authors propose, might be due to cyclin A’s entering the nucleus too soon. Recent fi ndings indicate that, although cyclin A is predominantly nuclear, it phosphorylates certain cytoplasmic proteins. Perhaps tethering cyclin A at the ER keeps it in the cytoplasm long enough for necessary S phase–promoting interactions. The nuclear allotment of cyclin A might be more important for G2 and M phase progression. Too much SCAPER would thus block mitosis by preventing its nuclear entry. Under normal conditions, cyclin A levels would be suffi ciently high by G2 that SCAPER can no longer restrain all of it. When communal cell death fails in mutant chimeric fl ies, epithelial tissue remains in patches (circled) across the wing.