Uncoupling of membrane ruffling and pinocytosis during Ras signal transduction.

Activation of Ras stimulates cell surface membrane ruffling and pinocytosis. Although seen as coupled events, our study demonstrates that membrane ruffling and pinocytosis are regulated by distinct Ras signal transduction pathways. Ras controls membrane ruffling via the small GTPase Rac. In BHK-21 cells, expression of the constitutively active Rac1(G12V) mutant, via a Sindbis virus vector, resulted in a dramatic stimulation of membrane ruffling without affecting the uptake of horseradish peroxidase. Expression of Ha-Ras(G12V), an activated Ras mutant, stimulated both membrane ruffling and horseradish peroxidase uptake. The Ha-Ras(G12V)-stimulated pinocytosis but not membrane ruffling was abolished by either wortmannin or co-expression with a dominant negative mutant of Rab5, Rab5(S34N). Expression of the activated Rab5(Q79L) mutant mimics the stimulatory effect of Ha-Ras(G12V) on pinocytosis but not membrane ruffling. Our data indicate that Ha-Ras(G12V) separately activates Rab5-dependent pinocytosis and Rac1-dependent membrane ruffling.

Parallel to the activation of nuclear gene expression, Ras-GTP also triggers profound changes in the cytoplasm. The Ras signaling cascade involving the small GTPase Rac1 plays an important role in regulating the actin cytoskeleton and cell surface membrane ruffles (20). Ras activation also stimulates pinocytosis (21) and transport to endosomes and lysosomes. This process involves the small GTPase, Rab5 (22,23). Because cell surface ruffling and pinocytosis both involve organized movements of the plasma membrane, and because Ras stimulates both processes, the two processes have long been thought to be linked (21). We have found that membrane ruffling and pinocytosis are regulated by distinct Ras signal transduction pathways. Whereas Ras-stimulated ruffling depends on Rac1 and is independent of PI 3-kinase and Rab5, Ras-stimulated pinocytosis depends on Rab5 and probably PI 3-kinase and is independent of Rac1.

MATERIALS AND METHODS
Cells and Wortmannin Treatment-BHK-21 cells were grown as described previously (23). Wortmannin (Sigma) in dimethyl sulfoxide at 1 mg/ml (stock solution) was freshly diluted with ␣-MEM and added to previously rinsed cell monolayers. The wortmannin treatment (15 min, 37°C) was followed by horseradish peroxidase (HRP) uptake. Dimethyl sulfoxide (Ͻ0.1% after dilution) had no effect on pinocytosis or membrane ruffling (data not shown).
Construction of Recombinant Sindbis Viruses-cDNAs of Ha-Ras, Rac1, and Rab5 were subcloned into the unique XbaI restriction site of the Sindbis vector Toto1000:3Ј2J (23,24). The plasmid was then linearized by XhoI digestion and used as a template for in vitro transcription with SP6 RNA polymerase. The resulting RNA transcripts were used for transfection of confluent BHK-21 cell monolayers using a Lipofectin-mediated procedure (Life Technologies, Inc.). Cells were maintained at 37°C, and the media containing released viruses were harvested 40 h after transfection. Virus titers were generally between 10 8 and 10 9 plaque-forming units per ml. Virus stocks were aliquoted and kept frozen at Ϫ80°C before use.
Immunoblot Analysis of Protein Expression-Cell lysates (5 l) were analyzed by SDS-polyacrylamide gel electrophoresis (12% acrylamide), and the proteins were transferred to an Immobilon-P membrane (Millipore) using a Bio-Rad semi-dry transfer apparatus. The membrane was probed with an Ha-Ras specific monoclonal antibody (AB1, from Oncogene Science), and the immunoblot was developed using the ECL reagents from Amersham Corp.
HRP Uptake Assay-Confluent BHK-21 monolayers were mock-infected or infected with the vector or recombinant viruses as described (23). At 4 h post-infection or otherwise indicated, cells were washed twice with serum-free ␣-MEM, and HRP uptake was initiated by adding HRP (5mg/ml) in ␣-MEM (1ml) containing 0.2% bovine serum albumin. Cell lysates were assayed for HRP activity as described (23).
Electron Microscopy of Early Endosomal Structures-At 4 h postinfection, early endosomes were marked by a 10-min HRP uptake. Cells were rinsed four times with ice-cold PBS, fixed in 2% glutaraldehyde (in PBS) for 30 min, and washed four more times with PBS. HRP reaction was developed for 10 min by incubating with 50 mM Tris-HCl, pH 7.4, containing 2 mg/ml diaminobenzidine (Sigma) and 0.01% H 2 O 2 (Sigma). Cells were then processed for Poly/Bed (Polysciences, Inc.) embedding, semi-thick section (ϳ200 nm) preparation, and electron microscopy.
Fluorescence Microscopy of Actin Architecture and Membrane Ruf-* This work was supported in part by grants from the National Institutes of Health (to P. D. S. and J. A. C.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
‡ fles-At 4 h post-infection, cells were fixed with 2% paraformaldehyde (in PBS) for 30 min, peameabilized, and quenched with PBS containing 0.05% Triton X-100, 0.1 N NH 4 Cl, and 0.2% gelatin, followed by staining with rhodamine-phalloidin (25). The coverslips were mounted in 70% glycerol (in PBS), and the actin localization and cell edge membrane ruffles were visualized by a Zeiss axiovert microscope and a Bio-Rad confocal scanning imaging system.

RESULTS AND DISCUSSION
Pinocytosis is regulated by serum growth factors but the molecular mechanisms and biological significance are not well understood. Many growth factors also stimulate cell surface membrane ruffling via rearrangement of actin filaments (20). It has long been thought that membrane ruffling contributes to increased pinocytosis (reviewed in Ref. 26). One example is oncogenic Ras-stimulated pinocytosis and membrane ruffling (21). To understand the molecular basis of the two processes, we used a combination of reagents that specifically alter pinocytosis and membrane ruffling and found, to our surprise, that pinocytosis and membrane ruffling are independent cellular processes regulated by distinct Ras signal transduction pathways. To determine the effect of Ras activation on pinocytosis, we transiently expressed a constitutively active Ras mutant, Ha-Ras(G12V), in cultured BHK-21 cells (23,24) and monitored pinocytosis of HRP (23). Fig. 1A shows immunoblot analysis of Ha-Ras(G12V) expression at different times after virus infection. At 4 h after infection, abundant Ha-Ras(G12V) expression was detected. Thereafter, the protein accumulated throughout virus infection (Fig. 1A). Endogenous Ras protein was not detected. HRP accumulation increased dramatically in cells expressing Ha-Ras(G12V) (Fig. 1B). At 4 h after infection, cells expressing Ha-Ras(G12V) showed a 3-fold stimulation of HRP uptake compared with the control cells (Fig. 1B). Ha-Ras(G12V) also increased the rate of HRP uptake (Fig. 1C). By electron microscopy, HRP was identified in intracellular endosomes (Fig. 2), which were significantly enlarged in cells expressing Ha-Ras(G12V) (Fig. 2), suggesting increased endosome activity.
Early endosome fusion immediately follows the internalization step at the plasma membrane and appears to be ratelimiting (22,23). Wortmannin inhibits a function required for the activation of Rab5 (33), a GTPase essential for early endosome fusion. Expression of the constitutively active Rab5 mutant, Rab5(Q79L), leads to increased early endosome fusion and increased pinocytosis while expression of the dominant negative Rab5(S34N) mutant results in decreased early endosome fusion and decreased pinocytosis (23,34). The positive correlation between early endosome fusion activity and cellular pinocytic activity led us to suggest that Ha-Ras(G12V) sequentially activates a wortmannin-sensitive function (possibly PI 3-kinase) and Rab5, leading to an increase in early endosome fusion and pinocytosis. We found that Rab5(Q79L) mimicked the stimulatory effect of Ha-Ras(G12V) on HRP uptake, and this stimulation could be blocked neither by wortmannin (33) nor by the dominant negative Ha-Ras(S17N) mutant (Fig. 3C). Co-expression of the activated Rab5(Q79L) and Ha-Ras(G12V) mutants did not lead to a synergistic stimulation of HRP uptake (Fig. 3C). Furthermore, like cells expressing Rab5(Q79L) (Fig. 2C), cells expressing Ha-Ras(G12V) exhibited enlarged endosomal structures (Fig. 2B), indicative of enhanced endosome fusion. Finally, co-expression of the dominant negative Rab5(S34N) mutant completely blocked the stimulatory effect of Ha-Ras(G12V) on HRP uptake (Fig. 3D).
The pioneering work of Hall and colleagues (20) has partially defined the Ras signal transduction pathway regulating membrane ruffling. Our study begins to shed light on this important   (20) previously characterized a Ras signal transduction pathway that promotes membrane ruffling via activation of the small GTPase Rac1. To further examine the relationship of Ras-stimulated membrane ruffling and pinocytosis, we expressed the constitutively active and dominant negative Rac1 mutants. As shown in Fig. 3E, expression of Rac1(G12V) did not stimulate HRP uptake while under the same conditions Ha-Ras(G12V) stimulated HRP uptake 3-fold. Fig. 3F shows the co-expression of constructs used in this study. These data immediately suggested that membrane ruffling and pinocytosis are controlled by distinct Ras signal transduction pathways. We then examined membrane ruffling (Fig. 4). Consistent with previous reports (20,25), expression of Rac1(G12V) resulted in a dramatic production of cell surface membrane ruffles (Fig.  4C). Under the same conditions, expression of Ha-Ras(G12V) also stimulated membrane ruffling (Fig. 4D). Cells expressing the activated Rab5(Q79L) mutant, a strong stimulator of pinocytosis, did not show membrane ruffling (Fig. 4E), nor did the vector virus-infected (Fig. 4B) or mock-infected control cells (Fig. 4A). In agreement with Nobes et al. (25), we found that wortmannin (100 nM) failed to block Rac(G12V)-induced or Ha-Ras(G12V)-induced membrane ruffling (data not shown), under conditions where Ha-Ras(G12V)-stimulated pinocytosis was completely inhibited (Fig. 3B). From these data, we conclude that the Ras signal transduction pathway leading to the activation of Rac1 and membrane ruffling does not contribute to the stimulation of pinocytosis.

CONCLUSIONS
Taken together, as summarized in Fig. 5, it is clear that pinocytosis and membrane ruffling are regulated by distinct sets of molecules during Ras signal transduction.