Proteins on the Move Endosomes Get a Tail

Intracellular bacteria, such as Listeria, use actin tails to propel themselves through the cytoplasm. Clearly the bacteria have highjacked the cell's actin-polymerization machinery. But, based on the observations of Taunton et al. that cells make their own actin-rich tails Taunton et al. (page 519), Listeria appear to have hijacked the idea not just of polymerization, but also of tail formation. Taunton et al. characterize endosomal vesicles that move in Xenopus oocytes and extracts using such tails.

At least in extracts, the vesicles are identified as endosomes based on their multivesicular structure and staining with acridine orange. Endosomal membranes from HeLa cells move when added to extracts, whereas plasma and endoplasmic reticulum membranes do not. Purified Golgi membranes may yet show motile activity, as Cdc42 has been shown to have a role in Golgi trafficking.

A function for the tails remains purely speculative. Possibilities include an ATP-dependent diffusion mechanism for organelles and a method for nucleating actin away from the plasma membrane.

Reducing the pH in endocytic compartments strips ligands from receptors, and activates pH-dependent proteases. But the function of the acidic environment of the trans-Golgi network (TGN) in exocytosis is not known. Henkel et al. (page 495) use a proton channel from influenza virus to equilibrate the TGN to cytoplasmic pH. (The channel normally helps acidify the contents of the virus in the endosome, thus dissociating viral proteins to allow subsequent nuclear import.) Under these conditions, the researchers show that the delivery of apical proteins in polarized cells is slowed.

Apical signal recognition is intact, as apical proteins are not redirected to the basolateral surface. And release of apical vesicles from the TGN is not altered, as the speed of apical delivery of a nonpolarized protein (human growth hormone) is not slowed. Thus the TGN machinery must be inhibited at the level of apical cargo incorporation into apical vesicles. Apical delivery may be mediated by incorporation of proteins into glycolipid-enriched rafts, so senior author Ora Weisz now plans to see if there are pH-dependent differences in the rafts before they are loaded into apical vesicles.

On page 427, Wilhelm et al. describe the isolation of a complex of eight polypeptides implicated in RNA localization in Drosophila. The complex was isolated using the protein Exuperantia (Exu) as a convenient handle, as Exu has been shown to be necessary for the localization of bicoid RNA at the anterior of the fly oocyte. Wilhelm et al. identify Ypsilon Schachtel (Yps) as the one protein in the complex that binds tightly to Exu. Yps contains a cold-shock domain, which in other systems has been implicated in translational regulation. Identification of the remaining six proteins should be made easier by the availability of the complete fly genome.

Wilhelm et al. look for RNAs in the complex and find not bicoid (possibly because of an instability element in the 3′ end of the bicoid RNA) but oskar. Exu and Yps proteins and oskar RNA share similar localization patterns, moving from the nurse cells to the anterior and then posterior of the oocyte. If, as expected, bicoid RNA is associated with the Exu complex, it must be shed before the complex moves to the posterior.

Flies lacking Exu show few or no defects in posterior development, so Exu RNA had not been previously suspected as an Exu cargo. Wilhelm et al. show, however, that accumulation of oskar RNA at the oocyte posterior is somewhat less efficient in flies lacking Exu. The remaining accumulation may occur through a combination of random mixing from cytoplasmic streaming and capture by anchoring proteins at the posterior.

The first gross sign of an axis in mouse embryos is the primitive streak, which forms on the posterior side of the embryo and extends anteriorly towards the node. Misexpression of a Wnt gene can cause anterior duplications of the axis. But more recent gene expression studies have shown that the anterior of the embryo is defined before the formation of the streak, and Huelsken et al. suggest that the Wnt pathway is also involved in this patterning event Huelsken et al. (page 567).

Huelsken et al. created a β-catenin knockout mouse, as this molecule is a common effector of many Wnt signaling molecules. Before the mutant embryos die, several anterior differentiation markers are either not expressed, or their expression does not shift to the prospective anterior. Chimeras demonstrate that β-catenin is required in the fetus-forming epiblast, but not the visceral endoderm, suggesting that β-catenin mediates a signal from the embryonic ectoderm to pattern the endoderm. Early anterior patterning may also involve cell migration. The consequences of Wnt signaling for these processes, and the targets of β-catenin signaling, remain unknown.

The first set of β-catenin knockout mice, created by other investigators, was not examined for anterior gene expression. The mice did, however, show disruption of cadherin-mediated cell adhesion. This defect may have arisen from a shortened, dominant-negative protein made from the first, incomplete knockout. In the new knockout Huelsken et al. do not see a defect in adhesion, but observe that plakoglobin is recruited to replace the adhesive function of β-catenin.

Multiplying cells increase in mass, but cells that increase in mass do not always divide. For example, whereas platelet-derived growth factor (PDGF)-BB causes the smooth muscle cells (SMCs) that line arteries to divide, angiotensin II (Ang II) results only in increased protein synthesis, or hypertrophy. PDGF-BB–dependent SMC proliferation is involved in formation of atherosclerotic plaques. SMC hypertrophy caused by Ang II may strengthen blood vessels, so that the vessels can withstand the greater strain induced by the vasoconstricting effects of Ang II.

Servant et al. have found the reason for these divergent responses Servant et al. (page 543). Both factors drive the accumulation of D-type cyclins and the activation of the partner Cdk4 kinase. But only PDGF-BB can fully activate the Cdk2 kinase needed for entry into S phase, because only PDGF-BB can convincingly turn off the transcription of the gene for p27Kip1, an inhibitor of Cdk2, and increase the turnover of the protein. The signal transduction pathways downstream of PDGF-BB and Ang II are at least partially overlapping, but analysis of the p27Kip1 promoter should help to determine the crucial differences.

By William A. Wells, 1095 Market St. #516, San Francisco, CA 94103. E-mail: