Advertisement
Signaling and Regulation

Geldanamycin Anisimycins Activate Rho and Stimulate Rho- and ROCK-Dependent Actin Stress Fiber Formation

Anahita Amiri, Farahnaz Noei, Tahir Feroz and Jonathan M. Lee
DOI: 10.1158/1541-7786.MCR-06-0362 Published September 2007

Abstract

Heat shock protein 90 (Hsp90) is a member of the heat shock family of molecular chaperones that regulate protein conformation and activity. Hsp90 regulates multiple cell signaling pathways by controlling the abundance and activity of several important protein kinases and cell cycle–related proteins. In this report, we show that inhibition of Hsp90 by geldanamycin or its derivative, 17-allylamino-17-desmethoxygeldamycin, leads to activation of the Rho GTPase and a dramatic increase in actin stress fiber formation in human tumor cell lines. Inactivation of Rho prevents geldanamycin-induced actin reorganization. Hsp90 inactivation does not alter the appearance of filopodia or lamellipodia and tubulin architecture is not visibly perturbed. Our observations suggest that Hsp90 has an important and specific role in regulating Rho activity and Rho-dependent actin cytoskeleton remodeling. (Mol Cancer Res 2007;5(9):933–42)

Keywords:
  • Geldanamycin
  • actin
  • Control of cell cycle progression
  • Cell motility and migration

Introduction

Heat shock proteins (Hsp) are molecular chaperones that promote correct protein folding and inhibit protein aggregation (1). Hsp proteins are classified into five families based on their molecular weights and functions: Hsp100, Hsp90, Hsp70, Hsp60, and the small Hsps. There are currently two identified human Hsp90 proteins, Hsp90 α and β. These two proteins exist as α-α and β-β homodimers and in complexes with other Hsp chaperones (2). Substrates for Hsp chaperone are described as clients and the association of Hsp90 with its client and with other Hsp proteins is regulated by a class of proteins termed cochaperones (3).

Hsp90 chaperone activity is dependent on the ATP occupancy of an NH2-terminal nucleotide-binding domain. It is believed that ATP regulates Hsp90 activity by causing it to move between two conformational states in two distinct cochaperone complexes. The current model is that Hsp90 first associates with its client in an ATP-unbound state in a multimer described as the “intermediate complex.” The intermediate complex consists of Hsp90 bound to client, the Hsp70 and Hsp40 chaperones, and the cochaperones Hip and Hop (3-6). ATP binds to Hsp90 in the intermediate complex, inducing a conformational change in Hsp90 that leads to Hsp70/Hsp40/Hip/Hop dissociation. The client remains bound to Hsp90. A second Hsp90 complex is then formed, the so-called “mature” complex. The mature complex consists of Hsp90, client and p23, Cdc37, and immunophilin cochaperones (4, 5). Within the mature complex, the client protein is held in an active conformational state and protected from proteolytic degradation. Subsequent hydrolysis of the Hsp90-bound ATP causes client release and dissociation of the mature complex components (7). Client release leads to its conformational inactivation and may also lead to its ubiquitin-mediated proteolytic degradation (8).

Hsp90 regulates the activity and abundance of close to 100 proteins (9). Most of these client proteins have a role in signal transduction and cell cycle control (9). Hsp90 substrates include several steroid receptors (10), the p53 tumor suppressor (11), the src tyrosine kinase (12), the Erb2/HER-2 receptor (13), and Akt/protein kinase B (14). A commonly used Hsp90 inhibitor is the Streptomyces metabolite geldanamycin (GA) or its derivative, 17-allylamino-17-desmethoxygeldamycin (17AAG; ref. 6). GA is a structural ATP mimic that acts as a competitive inhibitor of ATP binding to Hsp90 (15, 16). The binding of GA to Hsp90 shifts the activity of Hsp90 chaperone complexes so that they now promote client degradation. The ability of GA to cause client degradation has made it a valuable tool in identifying Hsp90-dependent signaling proteins. In addition, the large number of Hsp90 clients that promote cell proliferation has led to the idea that Hsp90 inactivation could be used as a cancer treatment (6). GA has in vitro antiproliferative capacities in multiple cancer cell lines (17) and in vivo antitumor effect (18). Currently, GA and its derivates are undergoing clinical trials for anticancer efficacy (6, 19).

In addition to regulating protein folding, Hsp proteins have a further role in regulating the cell cytoskeleton (19). Hsp70 binds tubulin and may directly or indirectly inhibit tubulin polymerization (20-22). Similarly, Hsp27 binds to actin and inhibits actin polymerization (23, 24). Hsp90 and Hsp100 have been reported to bind actin (25), but their role in regulating the actin cytoskeleton organization is unclear.

Here, we show that Hsp90 is an inhibitor of Rho activity. Rho is the founding member of the Rho family of GTPases (Rac, Cdc42, and Rho) that collectively regulate a variety of cellular functions, including organization of actin cytoskeleton (26). Rho induces assembly of contractile actin-myosin filaments through activation of serine/threonine ROCK kinase. ROCK increases the actin filament cross-linking activity of myosin II by phosphorylation-induced activation of myosin light chain as well as phosphorylation-induced inactivation of myosin light chain phosphatase (27). We find that Hsp90 inhibition by GA or 17AAG treatment increases the activity of Rho and results in induction of Rho/ROCK–dependent stress fiber formation. This suggests a role for Hsp90 in regulating Rho activity and the Rho-dependent actin remodeling.

Results

Hsp90 Inhibition Activates Stress Fiber Formation

It has previously been reported that Hsp90 is able to bind actin (25). To examine the role of Hsp90 in the actin cytoskeleton regulation, we used GA and 17AAG to inhibit Hsp90 function. As shown in Fig. 1A , Hsp90 inactivation by these two drugs causes cell cycle arrest. Twenty-four hours after treatment, the majority of 17AAG- and GA-treated cells are found in G2 (4N) populations. GA- and 17AAG-treated cells assume two distinct morphologies: either they retain the cuboidal shape of control cells or they assume a rounded shape (Fig. 1B). The rounded cells are arrested in metaphase, as evidenced by visible chromosomes aligned at the metaphase plate. The cuboidal cells have uncondensed chromosomes. The remainder of this article focuses on the cuboidal, nonmetaphase cells.

FIGURE 1.
FIGURE 1.

Hsp90 inactivation causes cell cycle arrest. A. HeLa cells treated for 24 h with either 17AAG or GA arrest in G2 phase of the cell cycle. B. Phase-contrast images of 17AAG- or GA-treated cells. Control cells have a cuboidal morphology and diffuse DNA staining indicative of interphase cells. Many drug-treated cells assume a rounded morphology (labeled R). These cells have condensed chromosomes aligned at the metaphase plate (labeled M). Bar, 30 μm.

To investigate whether Hsp90 inactivation has any effect on the actin cytoskeleton, we stained the GA- and 17AAG-treated cells with phalloidin. Untreated HeLa cells show filamentous actin (F-actin) at the cell periphery (lamellipodia) and in localized plaques likely to be focal adhesions (Fig. 2A ). Each cell has few prominent stress fibers and these fibers are generally found at the periphery of the cytoplasm. However, overnight treatment of cells with either 17AAG or GA leads to a dramatic increase of actin stress fiber formation (Fig. 2A). These new prominent stress fibers cross the length of the cell. There is no visible change in lamellipodia appearance. Stress fibers are visible as early as 2 h after GA treatment (Fig. 2B). In three independent experiments, ∼70% of GA-treated cells had prominent stress fibers compared with only 6% of control cells. Treatment of HeLa cells with a combination of GA and lysophosphatidic acid, a bioactive lysophospholipid that can affect cytoskeletal remodeling, is able to further increase the number of cells with stress fibers (70% to 99%; data not shown). Thus, lysophosphatidic acid and GA likely induce stress fibers through independent pathways. GA was also able to increase the number of stress fibers in SKOV3 cells compared with controls (Fig. 2C). SKOV3 is a human ovarian carcinoma cell line. Thus, the ability of Hsp90 inhibition to activate stress fiber formation is not cell line specific.

FIGURE 2.
FIGURE 2.

Hsp90 inactivation causes the appearance of stress fibers. A. HeLa cells treated for 16 h with either 17AAG or GA have visibly increased stress fibers relative to control DMSO-treated cells. Actin in control cells is found in lamellipodia (L) and focal adhesions (A). In treated cells, prominent stress fibers (S) appear. Lamellipodia appear unchanged. B. HeLa cells treated for 2 h with GA have visibly increased stress fibers relative to control DMSO-treated cells. C. SKOV3 cells treated for 2 h with GA have visibly increased stress fibers relative to control DMSO-treated cells. D. HeLa cells grown on poly-d-lysine–containing plates have a similar actin structure to cells grown on normal substrates except for prominent fillopodia-like protrusion (F) from the membrane. 17AAG and GA treatment increases stress fibers in cells grown on poly-d-lysine. Bar, 30 μm.

To determine whether the substrate matrix could modulate this increase in stress fiber appearance, we plated HeLa cells on poly-d-lysine–coated substrate and stained them with phalloidin. The poly-d-lysine provides a net positive charge that enhances cell/substrate adhesion. As shown in Fig. 2D, the actin distribution in HeLa cells grown on poly-d-lysine plates was similar to that of standard plates except for the presence of filopodia-like extensions from the cell lamellipod. GA and 17AAG treatment did not visibly affect these filopodia-like extensions, nor was lamellipodia appearance altered. However, as in the case of HeLa grown on uncoated cells, the appearance of stress fibers was visibly enhanced. This indicates that 17AAG and GA induce stress fibers independent of cell/substrate interaction.

Hsp90 Inactivation Does Not Affect Tubulin Structure

To determine whether Hsp90 inactivation might affect tubulin as well as actin structure, we stained GA- and 17AAG-treated HeLa cells with a β-tubulin antibody. As shown in Fig. 3 , the tubulin network in untreated HeLa cells is a fine meshwork of interconnected tubulin strands. The mitotic arrest induced by Hsp90 inactivation leads to the formation of prominent spindles in the rounded cells with metaphase chromosomes. However, the cuboidal cells that show increase in stress fibers do not substantially differ in tubulin structure relative to untreated HeLa cells. This suggests that the ability of Hsp90 inhibition to alter actin stress fiber formation is not the result of an overall disruption of the cell cytoskeleton.

FIGURE 3.
FIGURE 3.

Hsp90 inactivation does not affect the tubulin cytoskeleton. HeLa cells treated for 24 h with either 17AAG or GA have similar tubulin staining pattern relative to control cells. Bar, 30 μm.

Inhibition of Hsp90 by GA Is Sufficient to Activate Rho

Because of the importance of the Rho GTPase in activating stress fiber formation, we examined the ability of GA to alter Rho activity. As shown in Fig. 4A , treatment of cells with GA for 30 min or 2 h greatly increased Rho activity relative to vehicle-only control. Overall Rho levels were unchanged. Rho activity returns to baseline levels after overnight exposure to GA. To further extend this study, we used an ELISA-based assay (G-LISA, Cytoskeleton) to measure Rho activity. As shown in Fig. 4B, GA treatment reproducibly doubled Rho activity.

FIGURE 4.
FIGURE 4.

GA treatment causes Rho activation. A. HeLa cells were treated with GA or vehicle control for 30 min, 2 h, or 16 h. Cells were lysed and Rho activation assay was measured by immunoprecipitation. Proteins were resolved on SDS-PAGE gel and analyzed by immunoblotting with anti-Rho antibody. B. HeLa cells were with GA or vehicle control for 2 h. Cell lysates were subjected to G-Lisa Rho activation assay. Columns, mean of triplicate measurements; bars, SD.

Stress Fiber Production in Hsp90-Inactivated Cells Is Dependent on Rho and ROCK Activity

We next determined whether activation of Rho was necessary for GA- and 17AAG-induced stress fiber formation. To this end, we transfected HeLa cells with a green fluorescent protein (GFP)–tagged dominant-negative Rho (dn-Rho). As shown in Fig. 5A , transfection of dn-Rho substantially reduced the formation of stress fibers induced by GA and 17AAG. Dn-Rho did not visibly affect the actin cytoskeleton in untreated HeLa cells (Fig. 5). These results suggest that the actin reorganization induced by Hsp90 inactivation is dependent on Rho activity.

FIGURE 5.
FIGURE 5.

A. GA-induced stress fiber formation is attenuated by dn-Rho and ROCK inhibition. HeLa cells were transiently transfected with dn-Rho (Rho N19) and treated with GA/17AAG for 16 h (top) or DMSO (bottom). Actin was visualized with Alexa Fluor 546 phalloidin and Rho visualized with anti-Myc antibody (green). Blue, DNA. Bar, 30 μm (top) and 9 μm (bottom). B. ROCK activity is required for GA/17AAG–induced stress fiber formation. ROCK inhibitor (Y27632) was applied to GA- or 17AAG-treated cells (as above) overnight. C. HeLa cells were treated with GA for 2 or 16 h. Cell lysates were subjected to immunoblotting with anti-LIMK1.

One of the downstream targets of Rho is ROCK, a serine/threonine kinase involved in regulation of myosin light chain phosphorylation. To examine whether ROCK activation is required for GA-induced stress fiber formation, we inhibited ROCK activity using Y27632. As shown in Fig. 5B, treatment of cells with Y27632 alone (10 μmol/L) did not dramatically alter the shape of the cells, although the lamellipod had thinner appearance. However, addition of Y27632 visibly attenuated GA- and 17AAG-induced stress fiber formation. These results are consistent with the idea that Hsp90 inhibition activates stress fiber formation through Rho/ROCK signaling.

LIM kinase (LIMK) is one of the downstream effectors of Rho/ROCK pathway, which stabilizes actin filaments by phosphorylation and inactivation of the actin-depolymerizing factor/cofilin (28). Li et al. showed that in 293T cells, Hsp90 stabilizes LIMK1 by promoting its homodimerization and transphosphorylation. We did not detect substantial changes in LIMK activity on GA treatment (Fig. 5C). Therefore, it is likely that GA affects stress fiber formation independent of LIMK1.

Hsp90 Inactivation Has No Affect on Actin Polymerization

One of the mechanisms by which Rho affects stress fiber organization is by stimulating actin polymerization through the activation of downstream effectors such as mDia and profilin. To test whether Hsp90 inactivation could affect actin polymerization, we first used Western blotting to compare total actin content in GA/17AAG–treated and untreated cells. As shown in Fig. 6A , 17AAG and GA treatment does not substantially change total actin levels. Tubulin levels are similarly unaffected. To determine whether Hsp90 inactivation altered the cellular ratio of F-actin to G-actin, we used a qualitative assay to measure relative F-actin and G-actin levels based on the differential detergent extractability of F-actin and G-actin (29). As shown in Fig. 6B, the levels of detergent-soluble G-actin and detergent-insoluble F-actin are approximately equal in control HeLa cells. In three independent experiments, neither 17AAG nor GA treatment altered the amount of F-actin relative to G-actin (Fig. 6B). On the other hand, treatment of HeLa cells with jasplakinolide, an agent that promotes intracellular actin polymerization (30, 31), increases the amount of F-actin relative to G-actin compared with the vehicle-only control (Fig. 6C). The inability of 17AAG and GA to alter the F-actin/G-actin ratio or the total amount of actin indicates that Hsp90 inactivation increases neither actin abundance nor net polymerization. Rather, stress fiber assembly is likely the result of bundling of existing fibers rather than a net increase in actin polymerization.

FIGURE 6.
FIGURE 6.

Hsp90 inactivation does not affect actin abundance or polymerization. A. HeLa cells treated for 24 h with either 17AAG or GA have the same total amount of actin and tubulin, as measured by Western blotting, relative to untreated cells. B. The relative amount of F-actin and G-actin in HeLa cells does not change after 17AAG or GA treatment. C. The actin-polymerizing agent jasplakinolide causes an increase in the relative amount of F-actin relative to G-actin.

Hsp90 Inactivation Alters Focal Adhesion Appearance

In addition to its role in the assembly of stress fibers, Rho controls the formation of focal adhesions. Much of the F-actin in untreated HeLa cells is found in cytosolic plaques that have the appearance of focal adhesions (Fig. 7 ). A change in stress fiber content is often associated with changes in the content of focal adhesions (32). To determine whether Hsp90 inactivation could alter the appearance of focal adhesions, we stained GA-treated HeLa cells for vinculin. Vinculin is a structural component of focal adhesions (33) and is commonly used to identify these structures. As shown in Fig. 7, vinculin is found in a perinuclear punctuate pattern in the cytosol in untreated HeLa cells. Some vinculin is found at the cell periphery, organized into oval shapes that colocalize with actin (arrows). These structures represent focal adhesions. In untreated cells, focal adhesions are found exclusively in the lamellipod at the cell periphery. On the other hand, GA treatment induces the formation of focal adhesions that are found behind as well as within the lamellipod. Focal adhesions also appear somewhat bigger. These changes occur as early as 2 h after treatment and are sustained for at least 16 h (not shown). To determine whether changes in focal adhesions in GA-treated cells correspond with the changes in Rho activity, we transfected HeLa cells with dn-Rho-GFP or GFP control constructs and stained them with anti-vinculin antibody. No striking difference was seen between the GA-treated cells that express dn-Rho and control cells (Fig. 7B).

FIGURE 7.
FIGURE 7.

Hsp90 inactivation affects focal adhesion appearance. HeLa cells treated for 16 h with either 17AAG or GA have bigger vinculin staining pattern relative to control cells. Arrows, focal adhesions that are formed behind lamellipods. Bar, 30 μm. Top, magnified versions of the bottom images. Bar, 12 μm. B. Vinculin staining pattern in GA-treated cells is not dependent on Rho activity.

Hsp90 Inactivation Does Not Affect Cell Migration and Invasion

Hsp90 has been shown to be involved in many malignant phenotypes such as invasion and metastasis through its oncogenic clients. Moreover, activation of Rho and increased formation of stress fibers following GA or 17AAG treatment suggested that Hsp90 might have a direct role in cell migration and invasion processes. To test these possibilities, cell migration and invasion assays were done in HeLa cell using transwell chambers. Treatment of HeLa cells with either GA or GA with the ROCK inhibitor Y27632 had no effect on cell migration and invasion (Fig. 8 ). The above observations suggest that the GA- and 17AAG-induced increase in stress fiber formation without increase in number of focal adhesions would not result in efficient cell migration or invasion.

FIGURE 8.
FIGURE 8.

Hsp90 inactivation does not affect migration and invasion. A. HeLa cells were serum starved and treated with GA, GA plus Y27632, vehicle control, or vehicle control plus Y27632 overnight and placed in a transwell migration chamber. Top, representative field of cells that have migrated into the chamber. Bottom, columns, migration expressed as mean percentage of vector-only controls from triplicate independent experiments; bars, SD. B. HeLa cells were serum starved and treated with GA or vehicle control overnight. Cells were subjected to the invasion assay using Matrigel-coated transwells for ∼48 h. Top, representative field of cells that have invaded through the Matrigel. Bottom, columns, invasion expressed as mean percentage of vector-only controls of triplicate independent experiments with triplicate counts; bars, SD.

Discussion

In this report, we show that inactivation of Hsp90 by 17AAG and GA leads to activation of Rho and an increase in Rho/ROCK–dependent actin stress fiber formation in human HeLa cells. Hsp90 inhibition did not visibly alter lamellipod appearance, induce filopodia formation, or alter cell migration or invasion. Neither did Hsp90 inactivation alter the microtubule network. Our results indicate that Hsp90 has a specific role in regulating Rho/ROCK–dependent actin stress fiber creation.

Stress fibers are bundles of actin filaments that traverse the length of a cell. Stress fibers regulate cell shape by providing mechanical rigidity and help control cell migration and motility by connecting the cytoplasm to the growth matrix via focal adhesions. Rho regulates eukaryotic cell polarity, shape, and migration by controlling stress fibers and focal adhesion formation (34). Microinjection of active forms of RhoA rapidly induces stress fiber formation in mammalian cells (32, 35). Rho activation has been reported to increase the number of vinculin-containing focal adhesions (32). It is thought that a Rho-dependent increase in the number of focal complexes concomitantly increases stress fiber number by providing additional sites from which stress fibers could form (32). We find that Hsp90 inactivation induces the formation of focal adhesions behind the lamellipod. Whereas we detect changes in focal adhesion appearance, these changes do not seem to be dependent on Rho activity, and we do not detect any changes in cell mobility in the GA-treated HeLa cells.

Whereas GA and 17AAG alter stress fiber formation, we do not observe substantial changes in cell migration. Motility is a multistep process that includes protrusion, adhesion, and retraction (36). Successful migration requires coordination between all these steps. The inability of GA/17AAG treatment to visibly alter filopodia and lamellipodia may be the reason why migration rates do not change on Hsp90 inhibition.

Hsp27, Hsp25, and Hsp26 have all been reported to inhibit actin polymerization (23, 24, 37). However, GA/17AAG treatments do not alter the relative abundance of polymerized actin in HeLa cells, indicating that Hsp90 inactivation is not enhancing actin polymerization. Thus, the stress fibers we see on Hsp90 inhibition likely derive from preexisting actin polymers. Similarly, the ability of Rho to induce stress fiber formation seems to depend mostly on the organization of existing F-actin into fibers rather than on a substantial increase in actin polymerization or abundance (38).

Hsp90 can bind actin filaments and has been reported to have actin bundling activity (25). Because stress fiber appearance requires an overall increase in cellular actin bundling, this activity seems at odds with our observation that Hsp90 inhibition by 17AAG and GA induces stress fiber formation. Perhaps the endogenous actin bundling activity of Hsp90 is unrelated to its ATP occupancy and would therefore be immune to 17AAG and GA treatment. Therefore, 17AAG and GA could induce stress fiber formation independent of the actin bundling activity of Hsp90 but through an Hsp90 client protein. In our model, Hsp90 inhibits Rho activity. On inactivation of Hsp90 by 17AAG and GA treatment, Rho is activated, concomitantly leading to ROCK activation and stress fiber formation. The lack of LIMK activation on GA/17AAG treatment indicates that although LIMK is known to be activated By ROCK, Hsp90 inactivation is likely inducing stress fibers through a Rho- and ROCK-dependent, but LIMK-independent, pathway. We have been unable to identify the mechanism by which Hsp90 inhibition activates Rho. We have been unable to detect direct physical interaction between Hsp90 and Rho, suggesting that any Hsp90-dependent inhibition of Rho function is likely to be indirect. We speculate that Hsp90 maintains an inhibitor of Rho in an active state. In the presence of GA/17AAG, this inhibitor now assumes a nonfunctional conformation or is degraded, thus stimulating Rho activity and concomitant stress fiber formation.

Recent studies stress the importance of Hsp90 for tumor cell growth and/or survival. Hsp90 is 2- to 10-fold higher in tumor cells than in normal cells, and a growing number of its clients are key components of multiple signaling pathways, which are used by cancer cells. Therefore, using Hsp90 inhibitors for anticancer therapy has the advantage of targeting multiple oncogenic pathways at once. In fact, Hsp90 inhibitor GA has been shown to have in vitro antiproliferative capacities in multiple cancer cell lines (17) and 17AAG has antitumor activity in human breast, colon, and prostate xenografts. 17AAG has shown encouraging results in clinical trials. In this study, we identified GA and 17AAG as activators of Rho/ROCK signaling and actin stress fiber formation. The results of our study may provide further insight into the use of Hsp90 inhibitors in anticancer therapy.

Materials and Methods

Cell Lines, Drug Treatments, and Reagents

HeLa, a human cervical adenocarcinoma cancer cell line, was obtained from the American Type Culture Collection (CCL-2) and maintained in MEMα (Invitrogen) supplemented with 10% fetal bovine serum (Invitrogen) and 1% antibiotic-antimycotic (Invitrogen) at 37°C, 5% CO2 in a humidified incubator. SKOV3, a human ovarian carcinoma cell line, was obtained from American Type Culture Collection and maintained in McCoy's medium (Life Technologies, Inc.) supplemented with 10% fetal bovine serum. Cells were treated with 0.5 μmol/L GA (InvivoGen) and 17AAG (InvivoGen) in complete medium. For GA and 17AAG treatments lasting up to 4 h, HeLa cells were serum starved overnight. For 16-h drug treatments, cells were serum starved concomitant with GA or 17AAG addition. Jasplakinolide (Invitrogen) and the ROCK inhibitor Y27632 (Calbiochem) were used at a concentration of 1 and 10 μmol/L, respectively, in complete medium. Lysophosphatidic acid (18:1; 1-oleoyl-2-hydroxy-sn-glycero-3-phosphate) was obtained from Avanti Polar Lipids. pcDNA3-EGFP (Addgene plasmid 13031) and pcDNA3-EGFP-RhoA-T19N (Addgene plasmid 12967; ref. 39) plasmids were obtained from Addgene, Inc. Anti–phospho-LIMK1/2 was obtained from Upstate Cell Signaling.

Immunofluorescence Microscopy

HeLa cells were grown on glass coverslips. After 17AAG or GA treatment, cells were fixed with 3.7% formaldehyde for 1 h followed by 0.1% Triton X-100 permeabilization for 5 min. For actin immunofluorescence, fixed and permeabilized cells were incubated for 1 h with Alexa Fluor 546 phalloidin (Molecular Probes). After washing thrice in PBS, cells were incubated with Hoechst 33258 (Sigma) for 20 min to stain nuclei. For vinculin and tubulin staining, cells were blocked with 1% bovine serum albumin for 1 h followed by incubation with mouse monoclonal antibodies to vinculin (overnight) or tubulin (1 h; Sigma). Cells were washed thrice with PBS and incubated with F(ab′)2 goat anti-mouse immunoglobulin G (Molecular Probes) labeled with Alexa 610 (vinculin) or Alexa 488 (for tubulin). Cells were mounted on slides using Permount. Slides were analyzed with Leica DM IRE2 using 461-, 520-, and 635-nm filter sets for blue, green, and red, respectively. Images were acquired with Retiga 12 bit camera (Leica) and deconvoluted using Velocity software (Improvision).

Rho Activation Assay

HeLa cells were serum starved overnight and treated with GA or DMSO for 30 min to 2 h. Rho activity in a total cell lysate was measured using either an immunoprecipitation assay (Upstate) or a luminescence based G-Lisa Assay (Cytoskeleton) according to the manufacturer's instructions. Luminescence analysis for G-Lisa Assay was done with Kodak MI 2000 software.

Cell Migration and Invasion Assays

Cells were serum starved overnight and GA or GA with Y27632 was added to the medium. The top chambers of 6.5-mm Corning Costar transwells were loaded with 0.2 mL of cells (5 × 105 cells/mL) in serum-free medium. Complete medium (0.6 mL) was added to the bottom wells and cells were incubated at 37°C overnight. Cells were fixed and stained according to Rodrigues et al. (40). Cells on the top layer were removed and the images of the cells at the bottom of the membrane were captured with Canon camera and a Zeiss Axio Vert microscope. The mean values were obtained from three individual experiments using Excel Microsoft software. For cell invasion assay, cells were serum starved overnight. Twenty-four-well cell culture inserts (8-μm pore size; BD Biosciences) were loaded with 0.5 mL of cells (5 × 105 cells/mL) in serum-free medium. Complete medium (0.5 mL) was added to the bottom wells and cells were incubated at 37°C for 1 to 2 days. Cells were fixed, stained, and analyzed as above.

Actin Assays

For total actin preparation, HeLa cells were washed once with PBS, trypsinized, and lysed with radioimmunoprecipitation assay buffer [150 mmol/L NaCl, 1% NP40, 0.5% DOL, 0.1% SDS, 50 mmol/L Tris-Cl (pH 8.0)] supplemented with 1× protease inhibitor cocktail (Roche). Protein quantifications were carried out with a Bradford Assay (Bio-Rad Protein Assay, Bio-Rad) according to the manufacturer's directions. A total of 20 μg of protein were used per lane. Western blotting for actin and tubulin (Sigma) was used at a concentration of 1:1,500 and 1:1,000, respectively. For detergent extraction of F-actin, a subconfluent 100-mm plate of HeLa cells was fractionated as previously described (29). Twenty microliters of each extraction were loaded and blotted for actin content as described above.

Acknowledgments

We thank John Copeland, Sujeeve Jeganathan, and Nadine Wipere-Bergeron for helpful comments and critical reading of this manuscript.

Footnotes

  • Grant support: National Cancer Institute of Canada, with funds from the Canadian Cancer Society.

  • 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.

  • Note: A. Amiri and F. Noei contributed equally to this work.

    • Accepted June 4, 2007.
    • Received October 26, 2006.
    • Revision received May 11, 2007.

References

  1. ↵
    Hartl FU, Hayer-Hartl M. Molecular chaperones in the cytosol: from nascent chain to folded protein. Science 2002;295:1852–8.
    OpenUrlAbstract/FREE Full Text
  2. ↵
    Minami Y, Kimura Y, Kawasaki H, Suzuki K, Yahara I. The carboxy-terminal region of mammalian HSP90 is required for its dimerization and function in vivo. Mol Cell Biol 1994;14:1459–64.
    OpenUrlAbstract/FREE Full Text
  3. ↵
    Young JC, Agashe VR, Siegers K, Hartl FU. Pathways of chaperone-mediated protein folding in the cytosol. Nat Rev Mol Cell Biol 2004;5:781–91.
    OpenUrlCrossRefPubMed
  4. ↵
    Hernandez MP, Sullivan WP, Toft DO. The assembly and intermolecular properties of the hsp70-Hop-hsp90 molecular chaperone complex. J Biol Chem 2002;277:38294–304.
    OpenUrlAbstract/FREE Full Text
  5. ↵
    Scheibel T, Buchner J. The Hsp90 complex-a super-chaperone machine as a novel drug target. Biochem Pharmacol 1998;56:675–82.
    OpenUrlCrossRefPubMed
  6. ↵
    Isaacs JS, Xu W, Neckers L. Heat shock protein 90 as a molecular target for cancer therapeutics. Cancer Cell 2003;3:213–7.
    OpenUrlCrossRefPubMed
  7. ↵
    Young JC, Hartl FU. Polypeptide release by Hsp90 involves ATP hydrolysis and is enhanced by the co-chaperone p23. EMBO J 2000;19:5930–40.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    Connell P, Ballinger CA, Jiang J, et al. The co-chaperone CHIP regulates protein triage decisions mediated by heat-shock proteins. Nat Cell Biol 2001;3:93–6.
    OpenUrlCrossRefPubMed
  9. ↵
    Pratt WB, Toft DO. Regulation of signaling protein function and trafficking by the hsp90/hsp70-based chaperone machinery. Exp Biol Med (Maywood) 2003;228:111–33.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    Cheung J, Smith DF. Molecular chaperone interactions with steroid receptors: an update. Mol Endocrinol 2000;14:939–46.
    OpenUrlCrossRefPubMed
  11. ↵
    Muller L, Schaupp A, Walerych D, Wegele H, Buchner J. Hsp90 regulates the activity of wild type p53 under physiological and elevated temperatures. J Biol Chem 2004;279:48846–54.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    Xu Y, Lindquist S. Heat-shock protein hsp90 governs the activity of pp60v-src kinase. Proc Natl Acad Sci U S A 1993;90:7074–8.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    Tikhomirov O, Carpenter G. Geldanamycin induces ErbB-2 degradation by proteolytic fragmentation. J Biol Chem 2000;275:26625–31.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    Sato S, Fujita N, Tsuruo T. Modulation of Akt kinase activity by binding to Hsp90. Proc Natl Acad Sci U S A 2000;97:10832–7.
    OpenUrlAbstract/FREE Full Text
  15. ↵
    Prodromou C, Roe SM, O'Brien R, Ladbury JE, Piper PW, Pearl LH. Identification and structural characterization of the ATP/ADP-binding site in the Hsp90 molecular chaperone. Cell 1997;90:65–75.
    OpenUrlCrossRefPubMed
  16. ↵
    Stebbins CE, Russo AA, Schneider C, Rosen N, Hartl FU, Pavletich NP. Crystal structure of an Hsp90-geldanamycin complex: targeting of a protein chaperone by an antitumor agent. Cell 1997;89:239–50.
    OpenUrlCrossRefPubMed
  17. ↵
    Supko JG, Hickman RL, Grever MR, Malspeis L. Preclinical pharmacologic evaluation of geldanamycin as an antitumor agent. Cancer Chemother Pharmacol 1995;36:305–15.
    OpenUrlCrossRefPubMed
  18. ↵
    Sasaki K, Yasuda H, Onodera K. Growth inhibition of virus transformed cells in vitro and antitumor activity in vivo of geldanamycin and its derivatives. J Antibiot (Tokyo) 1979;32:849–51.
    OpenUrlCrossRefPubMed
  19. ↵
    Liang P, MacRae TH. Molecular chaperones and the cytoskeleton. J Cell Sci 1997;110:1431–40.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    Sanchez C, Padilla R, Paciucci R, Zabala JC, Avila J. Binding of heat-shock protein 70 (hsp70) to tubulin. Arch Biochem Biophys 1994;310:428–32.
    OpenUrlCrossRefPubMed
  21. Paradis D, Vallee F, Allard S, et al. Comparative study of pharmacokinetics and serum bactericidal activities of cefpirome, ceftazidime, ceftriaxone, imipenem, and ciprofloxacin. Antimicrob Agents Chemother 1992;36:2085–92.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    Kirby BA, Merril CR, Ghanbari H, Wallace WC. Heat shock proteins protect against stress-related phosphorylation of tau in neuronal PC12 cells that have acquired thermotolerance. J Neurosci 1994;14:5687–93.
    OpenUrlAbstract
  23. ↵
    Miron T, Vancompernolle K, Vandekerckhove J, Wilchek M, Geiger B. A 25-kD inhibitor of actin polymerization is a low molecular mass heat shock protein. J Cell Biol 1991;114:255–61.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    Benndorf R, Hayess K, Ryazantsev S, Wieske M, Behlke J, Lutsch G. Phosphorylation and supramolecular organization of murine small heat shock protein HSP25 abolish its actin polymerization-inhibiting activity. J Biol Chem 1994;269:20780–4.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    Koyasu S, Nishida E, Kadowaki T, et al. Two mammalian heat shock proteins, HSP90 and HSP100, are actin-binding proteins. Proc Natl Acad Sci U S A 1986;83:8054–8.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    Jaffe AB, Hall A. RHO GTPases: biochemistry and biology. Annu Rev Cell Dev Biol 2005;21:247–69.
    OpenUrlCrossRefPubMed
  27. ↵
    Riento K, Ridley AJ. Rocks: multifunctional kinases in cell behaviour. Nat Rev Mol Cell Biol 2003;4:446–56.
    OpenUrlCrossRefPubMed
  28. ↵
    Li R, Soosairajah J, Harari D, et al. Hsp90 increases LIM kinase activity by promoting its homo-dimerization. FASEB J 2006;20:1218–20.
    OpenUrlAbstract/FREE Full Text
  29. ↵
    Lyubimova A, Bershadsky AD, Ben-Ze'ev A. Autoregulation of actin synthesis responds to monomeric actin levels. J Cell Biochem 1997;65:469–78.
    OpenUrlCrossRefPubMed
  30. ↵
    Copeland JW, Treisman R. The diaphanous-related formin mDia1 controls serum response factor activity through its effects on actin polymerization. Mol Biol Cell 2002;13:4088–99.
    OpenUrlAbstract/FREE Full Text
  31. ↵
    Holzinger A. Jasplakinolide. An actin-specific reagent that promotes actin polymerization. Methods Mol Biol 2001;161:109–20.
    OpenUrlPubMed
  32. ↵
    Ridley AJ, Hall A. The small GTP-binding protein rho regulates the assembly of focal adhesions and actin stress fibers in response to growth factors. Cell 1992;70:389–99.
    OpenUrlCrossRefPubMed
  33. ↵
    Critchley DR. Focal adhesions—the cytoskeletal connection. Curr Opin Cell Biol 2000;12:133–9.
    OpenUrlCrossRefPubMed
  34. ↵
    Wherlock M, Mellor H. The Rho GTPase family: a Racs to Wrchs story. J Cell Sci 2002;115:239–40.
    OpenUrlFREE Full Text
  35. ↵
    Burridge K, Wennerberg K. Rho and Rac take center stage. Cell 2004;116:167–79.
    OpenUrlCrossRefPubMed
  36. ↵
    Heath JP, Holifield BF. Cell locomotion: new research tests old ideas on membrane and cytoskeletal flow. Cell Motil Cytoskeleton 1991;18:245–57.
    OpenUrlCrossRefPubMed
  37. ↵
    Rahman DR, Bentley NJ, Tuite MF. The Saccharomyces cerevisiae small heat shock protein Hsp26 inhibits actin polymerisation. Biochem Soc Trans 1995;23:77S.
    OpenUrlFREE Full Text
  38. ↵
    Machesky LM, Hall A. Rho: a connection between membrane receptor signalling and the cytoskeleton. Trends Cell Biol 1996;6:304–10.
    OpenUrlCrossRefPubMed
  39. ↵
    Subauste MC, Von Herrath M, Benard V, et al. Rho family proteins modulate rapid apoptosis induced by cytotoxic T lymphocytes and Fas. J Biol Chem 2000;275:9725–33.
    OpenUrlAbstract/FREE Full Text
  40. ↵
    Rodrigues SP, Fathers KE, Chan G, et al. CrkI and CrkII function as key signaling integrators for migration and invasion of cancer cells. Mol Cancer Res 2005;3:183–94.
    OpenUrlAbstract/FREE Full Text
Back to top
Molecular Cancer Research: 5 (9)
September 2007
Volume 5, Issue 9

Sign up for alerts

View this article with LENS

Open full page PDF
Article Alerts
Email Article
Citation Tools
Geldanamycin Anisimycins Activate Rho and Stimulate Rho- and ROCK-Dependent Actin Stress Fiber Formation
Anahita Amiri, Farahnaz Noei, Tahir Feroz and Jonathan M. Lee
Mol Cancer Res September 1 2007 (5) (9) 933-942; DOI: 10.1158/1541-7786.MCR-06-0362
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section

Advertisement