A novel role for inhibitor of apoptosis (IAP ... - The FASEB Journal

The FASEB Journal • Research Communication

A novel role for inhibitor of apoptosis (IAP) proteins as regulators of endothelial barrier function by mediating RhoA activation Michael C. Hornburger,*,‡,1 Bettina A. Mayer,‡,1 Stefanie Leonhardt,‡ Elisabeth A. Willer,‡ Stefan Zahler,‡ Andrea Beyerle,* Krishnaraj Rajalingam,† Angelika M. Vollmar,‡ and Robert Fürst*,2 *Institute of Pharmaceutical Biology, Biocenter, and †Institute of Biochemistry II, Medical School, Goethe University Frankfurt/Main, Frankfurt/Main, Germany; and ‡Pharmaceutical Biology, Department of Pharmacy, Center for Drug Research, University of Munich, Munich, Germany Inhibitor of apoptosis (IAP) proteins, such as XIAP or cIAP1/2, are important regulators of apoptosis in cancer cells, and IAP antagonists are currently evaluated as antitumor agents. Beyond their function in cancer cells, this study demonstrates a novel role of IAPs as regulators of vascular endothelial permeability. Two structurally different IAP antagonists, ABT and Smac085, as well as silencing of IAPs, reduced the thrombin receptor-activating peptide (TRAP)-induced barrier dysfunction in human endothelial cells as assessed by measuring macromolecular permeability or transendothelial electrical resistance. ABT diminished thrombin-evoked stress fiber formation, activation of myosin light chain 2, and disassembly of adherens junctions independent of calcium signaling, protein kinase C, and mitogen-activated protein kinases. Interestingly, ABT and silencing of IAPs, in particular XIAP, reduced the TRAP-evoked RhoA activation, whereas Rac1 was not affected. XIAP and, to a lesser extent, cIAP1 were found to directly interact with RhoA independently of the RhoA activation status. Under cell-free conditions, XIAP did not induce an ubiquitination of RhoA. In summary, our work discloses IAPs as crucial regulators of endothelial permeability and suggests IAP inhibition as interesting approach for the prevention of endothelial barrier dysfunction.—Hornburger, M. C., Mayer, B. A., Leonhardt, S., Willer, E. A., Zahler, S., Beyerle, A., Rajalingam, K., Vollmar, A. M., Fürst, R. A novel role for inhibitor of apoptosis (IAP) proteins as regulators of

ABSTRACT

Abbreviations: BIR, baculoviral iap repeat; cIAP, cellular inhibitor of apoptosis; EC, endothelial cell; FCS, fetal calf serum; GST, glutathione S-transeferase; HEK, human embryonic kidney; HMEC, human microvascular endothelial cell; HUVEC, human umbilical vein endothelial cell; IAP, inhibitor of apoptosis; ICAM-1, intercellular adhesion molecule 1; MAPK, mitogen-acitvated protein kinase; MLC, myosin light chain; MYPT, myosin phosphatase-targeting subunit; PAR, protease-activated receptor; PKC, protein kinase C; TEER, transendothelial electrical resistance; TRAP, thrombin receptor-activating peptide; XIAP, Xlinked inhibitor of apoptosis 1938

endothelial barrier function by mediating RhoA activation. FASEB J. 28, 1938 –1946 (2014). www.fasebj.org Key Words: small RhoA GTPases 䡠 inflammation 䡠 permeability Inhibitor of apoptosis (IAP) proteins are known inhibitors of caspases and, thus, suppressors of apoptosis. Overexpression of IAPs is found in various types of cancer and frequently correlates with high malignancy and poor prognosis. Consequently, IAP antagonists have been developed based on the structure of the endogenous IAP antagonist Smac and are currently tested in clinical trials as anticancer drugs (1). Cellular IAP1 and 2 (cIAP1/2) and X-linked IAP (XIAP) are the most important representatives of so far 8 described human IAPs, which are characterized by the presence of at least 1 baculoviral iap repeat (BIR) domain as common feature (2). In the last few years, a new function of IAPs as regulators of inflammatory signaling processes has emerged (3–5). The inflammatory response is a crucial part of the body’s immune defense and tissue repair program (6). Most of the current pharmacological regimes used to treat inflammatory diseases lack selectivity in regard to the disease-triggering pathological signaling events. Therefore, a deeper understanding of the mechanisms underlying the inflammatory response is still a great need to force antiinflammatory drug discovery (7). In this context, we could recently provide evidence that IAPs are very interesting anti-inflammatory drug targets, since they interfere with endothelial processes governing the interaction of neutrophils with the en1

These authors contributed equally to this work. Correspondence: Institute of Pharmaceutical Biology, Biocenter, Goethe University Frankfurt/Main, Max-von-LaueStr. 9, 60438 Frankfurt/Main, Germany. E-mail: fuerst@em. uni-frankfurt.de doi: 10.1096/fj.13-235754 This article includes supplemental data. Please visit http:// www.fasebj.org to obtain this information. 2

0892-6638/14/0028-1938 © FASEB

dothelium (3). Beyond this interaction as crucial part of the immune response, endothelial cells (ECs) also tightly regulate blood vessel permeability, a hallmark of inflammation. Typical proinflammatory stimuli, such as histamine or thrombin, are able to disrupt EC junctions and to evoke endothelial barrier dysfunction, i.e., vascular leakage and edema formation (8). Cytoskeleton rearrangements are initiated by e.g., the thrombintriggered activation of the protease-activated receptor (PAR), which is followed by an induction of downstream signaling events involving calcium, protein kinase C (PKC), small Rho GTPases and myosin light chain (MLC). In consequence, stress fibers are formed, the contractile machinery is activated, interendothelial junctions disassemble, and paracellular permeability is increased (9). We hypothesized that IAPs might play a role in these processes. Thus, we aimed to investigate the involvement of IAPs in the regulation of permeability, stress fiber formation, and RhoA activation in the human endothelium.

nate (FITC)-dextran (40 kDa, 1 mg/ml; Sigma-Aldrich, Taufkirchen, Germany) was used as macromolecular tracer. On treatment, samples were taken from the lower chamber at the indicated time points. Fluorescence (␭ex: 485 nm; ␭em: 535 nm) was measured with a fluorescence plate reader (SpectraFluor Plus; Tecan, Männedorf, Switzerland). The mean fluorescence of vehicle control-treated cells at the final time point was set as 1. Data are expressed as relative changes compared with control levels. Transendothelial electrical resistance (TEER) TEER was measured as described previously (13). Briefly, HUVECs were seeded onto collagen A (Biochrom)-coated Millicell 12-mm PCF inserts (Merck Millipore, Darmstadt, Germany). Inside a heated, custom-built, modified Ussing-like chamber, cells were kept in HEPES containing 10% FCS buffer, while bidirectional square current pulses of 50 ␮A (200 ms duration) were applied every third second. A custom-built voltage/current-clamp unit in connection with a computersupported evaluation program was utilized to calculate TEER in accordance with Ohm’s law from recorded resistance values. Cytosolic calcium imaging

MATERIALS AND METHODS IAP antagonists ABT (compound 11 reported in Oost et al., ref. 10) was a kind gift from Abbott Bioresearch (Worcester, MA, USA). The dimeric compound Smac085 (unpublished) was kindly provided by Pierfausto Seneci (Department of Organic and Industrial Chemistry, University of Milano, Milan, Italy). The functionality of Smac085 was shown by proving the downregulation of cIAP1 in ECs (data not shown). Cell culture ECs were cultured in EC growth medium (ECGM; PromoCell, Heidelberg, Germany) containing 10% heat-inactivated fetal calf serum (FCS), penicillin (100 U/ml)/streptomycin (100 ␮g/ml), and amphotericin B (2.5 ␮g/ml; PAA/GE, Pasching, Austria) at 37°C in 5% CO2. Primary human umbilical vein ECs (HUVECs) were freshly isolated from umbilical cords (obtained from local hospitals in accordance with the Declaration of Helsinki) by digestion with collagenase A (Roche, Mannheim, Germany). HUVECs were used until passage 3. The human microvascular EC line CDC/EU.HMEC-1 [human microvascular endothelial cells (HMECs)] was kindly provided by the U.S. Centers for Disease Control and Prevention (CDC; Atlanta, GA, USA) and used until passage 12. HMECs are proven to preserve morphology, phenotype, and functionality of normal human microvascular ECs (11). Human embryonic kidney (HEK) 293T cells were cultured in DMEM supplemented with 10% FCS (Gibco-BRL, Carlsbad, CA, USA) and penicillin (100 U/ml)/streptomycin (100 ␮g/ml) (Gibco-BRL) at 37°C in 5% CO2. Macromolecular permeability assay Macromolecular permeability was measured as described previously (12). Briefly, collagen G (Biochrom, Berlin, Germany)-coated Transwell plate inserts (pore size 0.4 ␮m, 12 mm diameter, polyester membrane; Corning, NY, USA) were used. HMECs were seeded into the upper chamber, grown to confluence and treated as indicated. Fluorescein isothiocyaIAPS REGULATE ENDOTHELIAL BARRIER FUNCTION

Intracellular calcium levels were measured as previously reported (14). In brief, HUVECs were grown to confluence on 8-well ␮-slides (Ibidi, Martinsried, Germany). Cells were loaded for 30 min with 2 ␮M Fura-2-AM (Biotrend, Cologne, Germany) in HEPES buffer and treated as indicated. Fluorescence measurements (␭ex: 340/380 nm; ␭em: 510 nm) were obtained by a Zeiss Axiovert 200 inverted microscope (⫻40; Zeiss, Göttingen, Germany) with a climatic chamber, a Polychrome V monochromator, and an IMAGO-QE camera (Till Photonics, Gräfelfing, Germany). A total period of 15 min with images being acquired every fifth second was analyzed with the TillVision 4.0.1.2 software (Till Photonics). Each data point of the different graphs was calculated from a randomly chosen rectangle containing ⱖ20 adjacent cells. Mean values are expressed. One representative plot of each graph is shown for clarity. Western blot analysis Western blot analysis was performed as described previously (14). The following antibodies were used: mouse anti-actin (C4; Merck Millipore), rabbit anti-phospho-(Ser) PKC substrate (Cell Signaling/New England Biolabs, Frankfurt/Main, Germany), rabbit anti-phospho-p38 mitogen-acitvated protein kinase (MAPK; T180/Y182; Cell Signaling), mouse anti-phosphop44/42 MAPK (ERK1/2; T202/Y204; Cell Signaling), rabbit anti-phospho-JNK (T183/Y185; Cell Signaling), mouse anti-VEcadherin (F-8; Santa Cruz Biotechnology, Heidelberg, Germany), rabbit anti-VE-cadherin (Y731; Invitrogen/Life Technologies, Darmstadt, Germany), rabbit anti-MLC2 (Santa Cruz Biotechnology), rabbit anti-phospho-MLC2 (T18/S19; Cell Signaling), mouse anti-phospho-MLC2 (S19), rabbit antimyosin phosphatase-targeting subunit 1 (MYPT1; Cell Signaling), rabbit anti-phospho-MYPT1 (T696; Merck Millipore), mouse anti-XIAP (BD Biosciences, Heidelberg, Germany), goat anti-human cIAP1 (R&D Systems, Wiesbaden, Germany), rabbit anti-HIAP1/BIR3 (cIAP2; Epitomics, Burlingame, CA, USA), rabbit anti-RhoA (Thermo Fisher Scientific, Dreieich, Germany), mouse anti-Rac1 (Thermo Fisher Scientific), mouse anti-c-myc (9E10; Santa Cruz), mouse anti-FLAG (Sigma-Aldrich), horseradish peroxidase (HRP)-conjugated donkey anti-goat (Santa Cruz Biotechnology), HRP-conjugated goat-anti-mouse (Santa Cruz Biotechnology or Biozol, 1939

Eching, Germany), HRP-conjugated goat anti-rabbit (Dianova, Hamburg, Germany; or Thermo Fisher Scientific), IRDye 680LT-conjugated goat anti-mouse (LI-COR Biosciences, Bad Homburg, Germany), and IRDye 800CW-conjugated goat anti-rabbit (LI-COR Biosciences). Immunocytochemistry and confocal laser scanning microscopy (CLSM) HUVECs were cultured in collagen G (Biochrom)-coated ␮-slides (Ibidi) until confluence. After treatment, cells were fixed with 10% formaldehyde (AppliChem, Darmstadt, Germany), permeabilized with 0.2% Triton X-100 (SigmaAldrich), and blocked with 0.2% BSA. The following antibodies/reagents were used: rabbit anti-phospho-MLC2 (T18/S19; Cell Signaling), rhodamine phalloidin (Invitrogen), and Alexa Fluor 488 goat anti-rabbit (Invitrogen). Images were acquired with a Zeiss LSM 510 META confocal microscope. RhoA and Rac1 affinity precipitation (pulldown assay) HUVECs were used for affinity precipitation of small GTPases by the Active RhoA or Rac1 Pull-Down and Detection kit (Thermo Fisher Scientific) according to the manufacturer’s instruction. Briefly, active RhoA (RhoA-GTP) and active Rac1 (Rac1-GTP) were captured from cell lysate by using glutathione S-transeferase (GST)-rhotekin RhoA or Rac1 binding domain (RBD)-fusion proteins immobilized on GST-agarose beads. Cell lysates were incubated with the fusion proteinloaded beads. By using spin columns, RhoA- or Rac1-GTP is eluted from the beads by a reducing sample buffer, which is further analyzed by Western blotting. Aliquots from the cell lysates are used for the detection of total protein contents of RhoA or Rac1 by Western blot analysis. Transfection of siRNAs and plasmids Transfection of HUVECs with siRNA was performed with the Amaxa HUVEC Nucleofector kit (Lonza, Cologne, Germany) or the Targefect-HUVEC kit (Targeting Systems, El Cajon, CA, USA) and with plasmids by using the Targefect-HUVEC kit. Cells were used for experiments 24 h after transfection. XIAP siRNA (On-TargetPlus SmartPool, Human BIRC4; Thermo Fisher Scientific), cIAP1 siRNA (On-TargetPlus SmartPool, Human BIRC2; Thermo Fisher Scientific) and nontargeting siRNA were from Dharmacon/Thermo Fisher Scientific. The following plasmids were obtained from Addgene (Cambridge, MA, USA): pcDNA3-EGFP (plasmid 13031) and pcDNA3-EGFP-RhoA wild-type (WT; plasmid 12965). HEK293T cells were cotransfected with pRK5-mycRhoA WT (plasmid 15899), pRK5-myc-RhoA Q63L (plasmid 15900), and pRK5-myc-RhoA T19N (plasmid 15901) with or without pEGZ-EV (empty vector) and pEGZ-FLAG-XIAP using GeneJuice transfection reagent (70967; Novagen, Merck Millipore) at a final concentration of 1 ␮g/ml.

beads (Sigma-Aldrich). For detecting the RhoA-XIAP interaction after overexpression, the cotransfected HEK293T cells were lysed 48 h post-transfection with lysis buffer as described previously (15). Myc-RhoA was then immunoprecipitated overnight using Myc antibody (see above). The antigenantibody complexes were precipitated by Sepharose-coupled protein A/G beads (Roche). In any case, the beads were washed with the lysis buffer and bound proteins were analyzed by SDS-PAGE and immunoblotting. GST pulldown assay GST fusion proteins of RhoA WT, RhoA Q63L, and RhoA T19N were purified following standard procedures. They were used for precipitation of recombinant XIAP as described previously (15). Briefly, GST proteins were incubated with glutathione Sepharose beads in binding buffer for 1 h at 4°C on a rotator. Beads were washed 3 times with binding buffer and pelleted and unspecific binding was blocked by incubation with 5% BSA containing binding buffer for 1–2 h. Beads were again washed 3 times and then incubated with purified XIAP (R&D Systems) at 4°C on a rotator for 1 h. Finally, the beads were washed with high-stringency binding buffer and boiled at 100°C for 5 min in 50 ␮l sample buffer. The interaction between the proteins was checked by SDS-PAGE as described above. Ubiquitination assay In vitro ubiquitination of RhoA and Rac1 was performed using RhoA and Rac1 WT protein purified from BL21-CodonPlus competent cells. Ubiquitination reaction was performed in the presence of ubiquitination buffer (50 mM Tris-HCl, pH 7.5; 100 mM NaCl; 2.5 mM MgCl2; and 1 mM DTT), 150 nM XIAP (R&D Systems), 100 nM E1 (Boston Biochem, Cambridge, MA, USA), 150 mM UbcH5a (Boston Biochem), 42 ␮M His-ubiquitin (Boston Biochem), 1⫻ Mg-ATP (Enzo Life Sciences), 1 U inorganic pyrophosphatase (Fluka; SigmaAldrich), and 50 mM DTT. The reaction was performed at 37°C for 30 min. For Western blot analysis, the reaction was stopped using Laemmli buffer and RhoA was visualized using RhoA antibody (26C4; Santa Cruz Biotechnology), while Rac1 was visualized using a Rac1 antibody (BD Biosciences). Statistical analysis Statistical analysis was performed by using Prism 5.04 software (GraphPad, San Diego, CA, USA). All experiments were independently performed ⱖ3 times. Bar graph data represent means ⫾ sem. Unpaired t test was used to compare 2 groups. Three or more groups were compared with 1-way ANOVA followed by Newman-Keuls post hoc test. Values of P ⱕ 0.05 was considered as statistically significant.

RESULTS Immunoprecipitation experiments HUVECs were used for immunoprecipitation of RhoA, cIAP1, and XIAP. Cells were either treated with ABT (1 ␮M) for 30 min, thrombin receptor-activating peptide (TRAP; 50 ␮M) for 10 min, or DMSO as vehicle control. Cell lysates were cleared by centrifugation for 15 min at 14,000 rpm at 4°C. cIAP1-RhoA, XIAP-RhoA, or RhoA-XIAP was then immunoprecipitated overnight using anti-RhoA, anti-XIAP, or anticIAP1 antibodies (see above). The antigen-antibody complexes were precipitated by agarose-coupled protein A/G 1940

Vol. 28

April 2014

IAP antagonists prevent TRAP-induced endothelial hyperpermeability Two complementary permeability assays were performed to determine the role of IAPs in the regulation of endothelial barrier function: measurement of macromolecular permeability and TEER. Pretreatment of a human EC monolayer with two structurally different IAP antagonists (ABT and Smac085) significantly re-

The FASEB Journal 䡠 www.fasebj.org

HORNBURGER ET AL.

duced the TRAP-induced flux of the macromolecular tracer compound FITC-dextran (40 kDa) through the endothelium (Fig. 1A–D): Fig. 1A, C shows the timedependent analysis of the dextran passage, and Fig. 1B, D shows the respective statistical analysis at time point 30 min. ABT (1 ␮M) or Smac085 (1 ␮M) inhibited the TRAP-induced endothelial hyperpermeability by 50%. TEER measurements allow the real-time analysis of ion flux across the endothelium and the calculated resistance values serve as readout parameter for the tightness of the barrier. As shown in Fig. 1E, pretreatment with ABT diminished the TRAP-induced drop of electrical resistance at any observed time point. Statistical evaluation of this effect (60 s after application of TRAP) is given in Fig. 1F: ABT attenuated the action of TRAP by 30%. Taken together, these results suggest that IAP antagonists protect endothelial barrier integrity, thus indicating a role for IAPs in the regulation of vascular permeability. IAP antagonism stabilizes adherens junctions and prevents cell contraction

IAP inhibition does not affect calcium levels nor PKC or MAPK signaling

Thrombin rapidly triggered the phosphorylation of VE-cadherin at Y731 within the first 5 min (Fig. 2A), which indicates adherens junction disassembly (16).

A

IAP inhibition by ABT strongly abolished this phosphorylation after 15 and 30 min (Fig. 2A). Regarding the actin cytoskeleton, the thrombin-induced stress fiber formation was completely reversed by ABT (Fig. 2B). The phosphorylation status of MLC2 controls the interaction of actin and myosin, which is indispensible for cell contraction (17). As shown in Fig. 2C, ABT reduced the thrombin-induced activation of the monophosphorylated (15 min) and diphosphorylated form (5 min) of MLC2 (at T18/S19) to control levels. MYPT1 is the regulatory subunit of the MLC phosphatase, which is responsible for the dephosphorylation of MLC. MYPT1 phosphorylation at T696 results in phosphatase inhibition (18) and thrombin treatment increased this phosphorylation, i.e., inhibited the phosphatase (Fig. 2D). ABT attenuated this phosphorylation at 5 and 15 min (Fig. 2D). These findings indicate that ABT protects against the disruption of adherens junctions and prevents stress fiber formation and activation of the contractile machinery.

Searching for the mechanisms underlying the barrierstabilizing effects, we investigated major endothelial

C

vehicle control

E 120

TRAP + ABT 0.1 µM TRAP + ABT 1 µM

2

1

0 0

10

20

30

vehicle control

3

ABT + TRAP

TRAP + Smac085 2

1

0

10

20

30

160 140

*

100

*

60 40 20 0

+

0.1 +

40

0

1 +

ABT [µM] TRAP

100

200

300

time [s]

F 200 120

180 160 140

TEER [%]

180

increase in FITC-dextran passage [%]

200

-

TRAP 40

time [min]

D

80

60

0 -100

0

40

B

120

80

20

time [min]


100

TRAP

TEER [%]

3

FITC-dextran passage [x-fold]

FITC-dextran passage [x-fold]

TRAP

120 100 80

*

60

100

80

60

40

*

20 0

-

+

+ +

Smac085 TRAP

40

-

+

+ +

ABT TRAP

Figure 1. IAP antagonists (ABT, Smac085) prevent TRAP-induced hyperpermeability. A–D) HMECs were pretreated for 30 min with ABT (1 ␮M; A, B) or Smac085 (1 ␮M; C, D) before TRAP (50 ␮M). Control cells were treated with DMSO as vehicle control. Flux of FITC-labeled dextran (40 kDa) across a human endothelial cell monolayer was measured with a Transwell 2-compartment system (A, C). Samples were taken from the lower compartment at the indicated time points. Bar diagrams show the statistical analysis at time point 30 min (B, D). E, F) TEER was measured in HUVECs with ABT pretreatment (1 ␮M) for 30 min before TRAP (50 ␮M). Control cells were treated with DMSO as vehicle control. E) Representative TEER curve. F) Statistical analysis of all TEER experiments (time point: maximum drop of resistance). Data are expressed as mean ⫾ sem of 3 independent experiments. *P ⬍ 0.05; 1-way ANOVA with Newman-Keuls post hoc test. IAPS REGULATE ENDOTHELIAL BARRIER FUNCTION

1941

C

A p-VE-cadherin

p-MLC2

VE-cadherin

pp-MLC2

β-actin -

+

-

+

-

+

-

+

ABT

-

-

5

5

15

15

30

30

thrombin [min]

MLC2 β-actin -

B

+ -

5

+ 5

15

+ 15

ABT thrombin [min]

D p-MYPT F-actin

MYPT

control

thrombin [5 min]

thrombin [5 min] + ABT

β-actin -

+

-

+

-

+

ABT

-

-

5

5

15

15

thrombin [min]

Figure 2. IAP antagonism stabilizes adherens junctions and prevents endothelial cell contraction. HUVECs were pretreated with ABT (1 ␮M) 30 min before thrombin (1 U/ml). Control cells were treated with DMSO as vehicle control. Phosphorylation of VE-cadherin at T731 (A), of MLC 2 at T18/S19 (C), and of the MYPT1 at T696 (D) was analyzed by Western blotting. Endothelial F-actin distribution was determined by confocal laser scanning fluorescence microscopy (B). Scale bar ⫽ 20 ␮m. Blots are representative of 3 independently performed experiments.

second messenger and signal transduction systems for an influence evoked by IAP inhibition. Intracellular Ca2⫹ concentrations ([Ca2⫹]i) were determined by ratiometric measurements using Fura-2-AM-loaded ECs. ABT did neither alter the rapid raise of [Ca2⫹]i caused by depletion of the endoplasmic reticulum, nor the long-lasting plateau phase that reflects the storeoperated calcium entry (Fig. 3A). The effect of ABT on the activity of PKC isoforms was investigated by analysis of the serine phosphorylation status of substrates with typical PKC phosphorylation motives. The induction of PKC activity was not altered by ABT treatment (Fig. 3B). Moreover, the potential of ABT to inhibit the thrombin-induced activation of MAPKs was analyzed at different time points: Neither the activation of ERK (measured by the phosphorylation status of T202/Y204, Fig. 3C), p38 MAPK (phospho-T180/Y182, Fig. 3D), or JNK (phospho-T183/Y185, Fig. 3E) was affected by ABT. Thrombin and ABT did not influence the expression of the levels of the pan proteins (ERK, p38 MAPK, and JNK; data not shown). In summary, these data exclude an influence of IAP inhibitors on endothelial calcium signaling, PKC, and the MAPK pathway. IAPs regulate RhoA activation and downstream events The action of IAP inhibition on the small Rho GTPases RhoA and Rac1, crucial regulators of the actin cytoskeleton, was investigated by pulldown assays. The thrombin-triggered RhoA activity was strongly reduced on 1942

Vol. 28

April 2014

treatment of ECs with ABT (Fig. 4A), whereas Rac1 was affected neither by thrombin nor by ABT (Fig. 4B). An influence of thrombin on the total protein levels of RhoA and Rac1 (Supplemental Fig. S1), as well as on the levels of XIAP and cIAP1 (Supplemental Fig. S2), was excluded. cIAP2 was found to be expressed neither in untreated nor in thrombin-activated ECs (Supplemental Fig. S3), thus ruling out an involvement of cIAP2. TNF-␣, which is known to increase cIAP2 protein levels (19), was used to demonstrate that cIAP2 can, in principle, be detected in our experimental setting (Supplemental Fig. S3). However, as shown in Fig. 4C, ABT rapidly depleted cIAP1 levels in human ECs, whereas XIAP levels were not affected. Thrombin did not influence this depletion. To investigate whether the presence of IAPs regulates RhoA activity, RNAi experiments were performed. Gene silencing of XIAP blocked activation of RhoA, and the knockdown of cIAP1 caused a partial attenuation (Fig. 4D). Of note, the silencing approach did not trigger a reciprocal up-regulation of the two IAPs (Fig. 4D). In addition, silencing of IAPs also blocked the TRAP-induced formation of stress fibers and activation of MLC2 and led to a more cortical actin distribution (Fig. 4E). Successful silencing was confirmed (Supplemental Fig. S4). Inhibition of stress fiber formation by ABT was also observed in a RhoA-overexpression system (Supplemental Fig. S5). Most notably, macromolecular permeability induced by TRAP was abolished in IAP-depleted HUVECs (Fig. 4F). These data point to an involvement of XIAP and cIAP1 in the thrombin/TRAP-induced activa-


HORNBURGER ET AL.

A

B

C

2.5

thrombin thrombin + ABT

F340/F380

2.0

p-ERK

p-PKC substrates

1.5

β-actin

1.0 0.5

β-actin

0 0

2

4

6

8

10

12

14

16

-

+ -

5

+ 5

- + + 15 15 30 30

-

+ -

5

+ 5

+ ABT 15 15 thrombin [min]

ABT thrombin [min]

Figure 3. IAP inhibition does not affect calcium levels or PKC or MAPK signaling. A) Change of [Ca2⫹]i was measured in Fura-2-AM-loaded D E HUVECs. Cells were pretreated for 30 min with ABT (1 ␮M) or with p-p38 p-JNK DMSO as vehicle control. At 1 min after starting the measurement, β-actin β-actin thrombin (1 U/ml) was added. B–E) + - + - + ABT HUVECs were pretreated with ABT + - + - + ABT (1 ␮M) for 30 min before thrombin 5 5 15 15 thrombin 5 5 15 15 thrombin [min] (1 U/ml). Control cells were treated [min] with DMSO as vehicle control. Phosphorylation status of PKC substrates (B), EKR (C), p38 MAPK (D), and JNK (E) was analyzed by Western blotting. Curves and blots are representative of 3 independently performed experiments. time [min]

tion of RhoA, F-actin, MLC, and endothelial barrier function. XIAP directly interacts with RhoA To test the association of IAPs with RhoA, immunoprecipitation experiments were performed in ECs. The association of cIAP with RhoA was detectable, but weak (Fig. 5A). In contrast, XIAP was found to strongly bind to RhoA (Fig. 5B). The binding of XIAP to RhoA seems not to depend on the activation status of RhoA, because TRAP treatment, which increases RhoA-GTP levels, did not affect the RhoA-XIAP interaction in ECs (Fig. 5C). The association of XIAP with RhoA was also confirmed in the reciprocal precipitation setting (Fig. 5D). Next, we aimed to prove that the RhoA activation status does not alter the interaction between XIAP and RhoA by using a constitutively active RhoA mutant (Q63L). HEK 293T cells were cotransfected with plasmids coding for XIAP and either the WT form of RhoA or the RhoA mutant Q63L. As shown in Fig. 5E, the association of RhoA with XIAP did not depend on the RhoA activation status. Moreover, in a cell-free system using recombinantly expressed, purified GST fusion proteins, the RhoA WT form, the RhoA mutant Q63L, and the dominant negative RhoA mutant T19N were found to directly interact with recombinant XIAP (Fig. 5F). Since IAPs can function as E3 ubiquitin ligases (20, 21), we tested whether the interaction of XIAP and RhoA affects ubiquitination processes. As shown in Supplemental Fig. S6, RhoA was not ubiquitinated in the presence of XIAP. Rac1 served as positive control, because XIAP is known to induce its ubiquitination (15). Taken together, these results suggest that IAPs directly interact with RhoA. This interaction does not depend on the GDP/GTP status of RhoA and RhoA is not ubiquitinated by XIAP. IAPS REGULATE ENDOTHELIAL BARRIER FUNCTION

DISCUSSION In this study, we demonstrate for the first time that IAPs are crucial players in the regulation of permeability and RhoA activation in response to thrombin in the human endothelium. We showed that IAP antagonism or gene silencing prevents thrombin-induced endothelial hyperpermeability, adherens junction disassembly, stress fiber formation, and contraction independent of the calcium, PKC or MAPK signaling pathway. Moreover, IAPs were found to mediate RhoA but not Rac1 activation. IAPs, in particular XIAP, were found to strongly interact with RhoA. This interaction is independent of the RhoA activation status, and RhoA is not ubiquitinated by XIAP. Beyond the role of IAPs in cancer cells, a novel function of IAPs as mediators of inflammatory signaling processes has emerged in the past years: IAPs were discovered as downstream mediators of important pattern recognition receptors, such as Toll-like receptor 4, NOD-like receptors, or RIG-I, and also of the TNF receptor-triggered NF-␬B and MAPK signaling pathways (3, 22–24). Surprisingly, although the endothelium is one of the most important cell types in the regulation of the inflammatory response, knowledge about the role of IAPs in the endothelium is very limited. Kearney et al. (5) showed that an IAP antagonist attenuates the TNF-␣-induced cytokine/ chemokine expression in HUVECs. Recently, we have provided further evidence that IAPs are actually involved in inflammation-associated endothelial processes: we have reported that IAP inhibition causes a reduction in intercellular adhesion molecule 1 (ICAM-1) expression in TNF-␣-activated ECs and prevents leukocyte recruitment to the site of inflammation in vivo by altering TNF receptor signaling (3). Besides TNF-␣, the serine protease thrombin is also a crucial proinflammatory mediator. It represents a procoagulant protein and is, for instance, found at in1943

A

D

B

-

+

+ -

+ +

C

RhoA-GTP

Rac1-GTP

XIAP

RhoA

Rac1

cIAP1

β-actin

β-actin

thrombin ABT

E

-

+

+ -

+ +

thrombin ABT

β-actin -

+ -

5

- + - + ABT 15 15 30 30 thrombin [min]

F-actin

RhoA-GTP RhoA

+ 5

cIAP1 XIAP + + -

+ + -

+ +

β-actin TRAP nt siRNA cIAP1 siRNA XIAP siRNA


F

pp-MLC2

+ -

nt siRNA 150

nt siRNA + TRAP

cIAP1 siRNA + TRAP

XIAP siRNA + TRAP

Figure 4. IAPs regulate RhoA activation and stress fiber formation. A–C) HUVECs were pretreated with ABT (1 ␮M) for 30 min before thrombin (1 U/ml, 10 min). Control cells were treated with DMSO as vehicle control. A, B) Active RhoA (RhoA-GTP; A) or active Rac1 (Rac1-GTP; B) were precipitated 50 from cell lysates (affinity precipitation/pulldown assay followed by Western blot analysis). Total RhoA/Rac1 levels were detected by Western blotting. C) XIAP and cIAP1 protein levels were detected by Western blotting. D–F) XIAP and 0 cIAP1 were down-regulated in HUVECs by siRNA. Nontargeting (nt) siRNA cIAP1 served as negative control. TRAP (50 ␮M) was applied for 10 min. D) Active RhoA (RhoA-GTP) was precipitated from cell lysates (affinity precipitation/ XIAP pulldown assay followed by Western blot analysis). Total RhoA, cIAP1, and XIAP β-actin levels were detected by Western blotting. E) Distribution of F-actin and phos+ + + TRAP phorylated MLC2 was determined by confocal laser scanning fluorescence + + nt siRNA microscopy out of the experiments performed in D. Scale bar ⫽ 20 ␮m. Images + cIAP1 siRNA are representative of ⱖ3 independent experiments. F) Flux of FITC-labeled + XIAP siRNA dextran (40 kDa) across the human endothelial cell monolayer was measured with a Transwell 2-compartment system. Samples were taken from the lower compartment over a time period of 30 min. RhoA, cIAP1, and XIAP levels were detected by Western blotting; successful silencing was proven by Western blot analysis using cIAP1 and XIAP antibodies. Bar diagram shows statistical analysis at time point 30 min. Blots are representative of ⱖ3 independent experiments. Data are expressed as means ⫾ sem. *P ⬍ 0.05; 1-way ANOVA with Newman-Keuls post hoc test. 100

*

*

*

creased levels in different lung disorders, such as acute lung injury or acute respiratory distress syndrome (25, 26). Thrombin is known to cause massive endothelial barrier dysfunction (27, 28) by activating PARs. Here, we have shown for the first time that IAPs are involved in the thrombin-induced inflammatory response by mediating endothelial RhoA activation, which triggers important RhoA downstream events. In the present study, we could demonstrate that the thrombin-triggered RhoA activation depends on the presence of both IAPs, with more pronounced effects observed on XIAP silencing. However, cIAP1, which is rapidly degraded on addition of an IAP antagonist by the proteasome, is more relevant for the TNF-induced ICAM-1 expression than XIAP (3). This suggests that IAP family members are not fully exchangeable and might exert distinct functions in the endothelium. In contrast to cIAP1, XIAP levels are not altered by the IAP inhibitor ABT. Originally, the Smac mimetic ABT was designed and proven to potently inhibit XIAP via association with the BIR3 domain (Kd: 16 nM; ref. 10). Of note, the precise 1944

Vol. 28

April 2014

mechanism of IAP antagonist-induced autoubiquitination and subsequent degradation of cIAP1 has recently been described (20). To date, there are first hints available about structural alterations induced by the binding of Smac mimetics to XIAP (21, 29): it is now clear that the RING domain of XIAP is not mainly addressed by Smac mimetics. Further investigations to pinpoint the exact binding domain of the respective IAP are needed for a better understanding of the underlying mechanism. Interestingly, we could prove a direct interaction of XIAP and RhoA, which is independent of the RhoA activation status under endogenous conditions and in vitro. A direct association of IAPs with Rho GTPases has as yet only been reported for Rac1 (30): in fruit flies, the IAP homologue DIAP1 was shown to affect cell migration by binding Rac1 in a nucleotide-independent manner (31). Furthermore, it was reported for mouse embryonic fibroblasts and HeLa cells that cIAP1 and XIAP interact with GTP- and GDP-loaded Rac1 (15). Oberoi et al. (15) found that IAPs directly target Rac1, trigger its degradation, and, consequently, lead to an increased Rac1 activation. In


HORNBURGER ET AL.

B

β-actin -

+

RhoA

-

+

ABT FLAG

RhoA-IP

XIAP-IP

XIAP

F

+

TRAP

XIAP

lysates

lysates

-

β-actin

myc

RhoA β-actin

*

β-actin

D RhoA

myc-IP

XIAP RhoA

ABT

C

RhoA

lysates

lysates

RhoA

293T

* denotes endogenous XIAP

XIAP-IP

RhoA

E

lysates

cIAP1-IP

A

β-actin -

+

ABT

+

-

+

-

empty vector

+

+

-

-

myc-RhoA wt

-

-

+

+

myc-RhoA Q63L

-

+

-

+

FLAG-XIAP

GST-RhoA T19N

GST-RhoA Q63L

GST-RhoA wt

GST

+ XIAP

XIAP

Figure 5. XIAP directly interacts with RhoA. A–D) HUVECs were treated with ABT (1 ␮M) for 30 min (A, B, D). HUVECs were treated with TRAP (50 ␮M) for 10 min (C). Control cells were treated with DMSO as vehicle control. Immunoprecipitation of cIAP1 (A), XIAP (B, C), or RhoA (D) was performed. RhoA (A–C) and XIAP (D) levels were determined by Western blotting. E) HEK293T cells were cotransfected with plasmids coding for myc-RhoA WT and myc-RhoA Q63L (constitutively active), with or without FLAG-XIAP. Immunoprecipitation of the myc-tag was performed. Levels of XIAP and of the myc- and FLAG-tag were analyzed by Western blotting. F) In a cell-free system, recombinant GST-RhoA WT, GST-RhoA Q63L and GST-RhoA T19N (dominant negative) were incubated with recombinant XIAP. GST pulldown experiments were performed. The presence of XIAP in the precipitate was analyzed by Western blotting. Blots are representative blot of 3 independently performed experiments.

Ponceau S

contrast to these studies, Rac1 was not influenced in our system. Of note, Oberoi et al. (15) showed that RhoA activity is influenced by IAPs, since the basal activity of RhoA was blocked by IAP antagonism. However, this effect was not further evaluated. Regarding the question of which domain of XIAP is responsible for the interaction with RhoA, we could show in preliminary experiments that a deletion of the RING domain severely impairs the direct interaction between XIAP and RhoA. Whether the RING domain directly mediates the interaction with RhoA deserves further investigations. One possibility is that the deletion of the RING domain compromises the conformational competence of XIAP in binding to RhoA. The RING domain of XIAP exhibits an E3 ubiquitin ligase function. Smurf1, a HECT domain E3 ligase, was recently reported to bind to RhoA and to regulate RhoA/RhoA kinase (ROCK)/MLC2 signaling (32). Thus, we speculated that ubiquitination processes might be involved in the regulation of RhoA activity and investigated the influence of XIAP on the ubiquitination of RhoA. In contrast to Rac1, which is known to be readily ubiquitinated by XIAP IAPS REGULATE ENDOTHELIAL BARRIER FUNCTION

(15), we could not detect any ubiquitination of RhoA. This points toward a different regulation of RhoA and Rac1 by XIAP and suggests that the E3 ligase function of XIAP might not play a direct role in the regulation of RhoA. The observed inhibition of RhoA activation by ABT or IAP silencing might involve typical RhoA upstream activators, such as GTPase-activating proteins (GAPs), guanine nucleotide exchange factors (GEFs), or guanine nucleotide dissociation inhibitors (GDIs). To date, only an interaction of XIAP with RhoGDI was demonstrated, which affects actin polymerization and motility of cancer cells (33). Preliminary data from our group showed that a RhoA-GEF (hDbs) is not influenced in its RhoA-activating capacity in the presence of XIAP, suggesting that GEFs are not involved in the XIAP-induced activation of RhoA. In summary, our study introduces IAPs, in particular XIAP, as crucial regulators of endothelial permeability by controlling the activity of RhoA. In consequence, IAP inhibition might be suggested as a novel approach to prevent endothelial barrier dysfunction. 1945

The compound A-4.10099.1 (ABT) was a kind gift of Abbott Bioresearch (Worcester, MA, USA). Smac085 was kindly provided by Pierfausto Seneci (Department of Organic and Industrial Chemistry, University of Milano, Milan, Italy). The authors thank Tripat K. Oberoi-Khanuja and Arun Murali (K.R. laboratory) for conceiving, performing, analyzing, and interpreting the experiments with the RhoA/XIAP mutants and on Rho/Rac ubiquitination. The excellent technical help of Jana Peliskova, Bianca Hager, and Silvia Schnegg is gratefully acknowledged. K.R. is a PLUS3 fellow of the Boehringer Ingelheim Foundation.

16.

REFERENCES

19.

1. 2. 3.

4.

5.

6. 7. 8.

9. 10.

11.

12.

13.

14.

1946

Fulda, S., and Vucic, D. (2012) Targeting IAP proteins for therapeutic intervention in cancer. Nat. Rev. Drug Disc. 11, 109 –124 Srinivasula, S. M., and Ashwell, J. D. (2008) IAPs: what’s in a name? Mol. Cell 30, 123–135 Mayer, B. A., Rehberg, M., Erhardt, A., Wolf, A., Reichel, C. A., Kracht, M., Krombach, F., Tiegs, G., Zahler, S., Vollmar, A. M., and Furst, R. (2011) Inhibitor of apoptosis proteins as novel targets in inflammatory processes. Arterioscler. Thromb. Vasc. Biol. 31, 2240 –2250 Varfolomeev, E., Goncharov, T., Maecker, H., Zobel, K., Komuves, L. G., Deshayes, K., and Vucic, D. (2012) Cellular inhibitors of apoptosis are global regulators of NF-kappaB and MAPK activation by members of the TNF family of receptors. Sci. Signal. 5, ra22 Kearney, C. J., Sheridan, C., Cullen, S. P., Tynan, G. A., Logue, S. E., Afonina, I. S., Vucic, D., Lavelle, E. C., and Martin, S. J. (2013) Inhibitor of apoptosis proteins (IAPs) and their antagonists regulate spontaneous and tumor necrosis factor (TNF)induced proinflammatory cytokine and chemokine production. J. Biol. Chem. 288, 4878 –4890 Medzhitov, R. (2008) Origin and physiological roles of inflammation. Nature 454, 428 –435 Foley, J. F. (2013) Focus issue: understanding mechanisms of inflammation. Sci. Signal. 6, eg2 Kumar, P., Shen, Q., Pivetti, C. D., Lee, E. S., Wu, M. H., and Yuan, S. Y. (2009) Molecular mechanisms of endothelial hyperpermeability: implications in inflammation. Exp. Rev. Mol. Med. 11, e19 Vandenbroucke, E., Mehta, D., Minshall, R., and Malik, A. B. (2008) Regulation of endothelial junctional permeability. Ann. N. Y. Acad. Sci. 1123, 134 –145 Oost, T. K., Sun, C., Armstrong, R. C., Al-Assaad, A. S., Betz, S. F., Deckwerth, T. L., Ding, H., Elmore, S. W., Meadows, R. P., Olejniczak, E. T., Oleksijew, A., Oltersdorf, T., Rosenberg, S. H., Shoemaker, A. R., Tomaselli, K. J., Zou, H., and Fesik, S. W. (2004) Discovery of potent antagonists of the antiapoptotic protein XIAP for the treatment of cancer. J. Med. Chem. 47, 4417–4426 Ades, E. W., Candal, F. J., Swerlick, R. A., George, V. G., Summers, S., Bosse, D. C., and Lawley, T. J. (1992) HMEC-1: establishment of an immortalized human microvascular endothelial cell line. J. Invest. Dermatol. 99, 683–690 Bubik, M. F., Willer, E. A., Bihari, P., Jurgenliemk, G., Ammer, H., Krombach, F., Zahler, S., Vollmar, A. M., and Furst, R. (2012) A novel approach to prevent endothelial hyperpermeability: the Crataegus extract WS(R) 1442 targets the cAMP/ Rap1 pathway. J. Mol. Cell. Cardiol. 52, 196 –205 Furst, R., Bubik, M. F., Bihari, P., Mayer, B. A., Khandoga, A. G., Hoffmann, F., Rehberg, M., Krombach, F., Zahler, S., and Vollmar, A. M. (2008) Atrial natriuretic peptide protects against histamine-induced endothelial barrier dysfunction in vivo. Mol. Pharmacol. 74, 1–8 Willer, E. A., Malli, R., Bondarenko, A. I., Zahler, S., Vollmar, A. M., Graier, W. F., and Furst, R. (2012) The vascular barrierprotecting hawthorn extract WS(R) 1442 raises endothelial calcium levels by inhibition of SERCA and activation of the IP3 pathway. J. Mol. Cell. Cardiol. 53, 567–577

Vol. 28

April 2014

15.

17. 18.

20.

21. 22. 23. 24.

25.

26. 27. 28.

29.

30. 31. 32.

33.

Oberoi, T. K., Dogan, T., Hocking, J. C., Scholz, R. P., Mooz, J., Anderson, C. L., Karreman, C., Meyer zu Heringdorf, D., Schmidt, G., Ruonala, M., Namikawa, K., Harms, G. S., Carpy, A., Macek, B., Koster, R. W., and Rajalingam, K. (2012) IAPs regulate the plasticity of cell migration by directly targeting Rac1 for degradation. EMBO J. 31, 14 –28 Lampugnani, M. G., and Dejana, E. (2007) Adherens junctions in endothelial cells regulate vessel maintenance and angiogenesis. Thromb. Res. 120(Suppl. 2), S1–S6 Shen, Q., Rigor, R. R., Pivetti, C. D., Wu, M. H., and Yuan, S. Y. (2010) Myosin light chain kinase in microvascular endothelial barrier function. Cardiovasc. Res. 87, 272–280 Kim, K. M., Csortos, C., Czikora, I., Fulton, D., Umapathy, N. S., Olah, G., and Verin, A. D. (2012) Molecular characterization of myosin phosphatase in endothelium. J. Cell. Physiol. 227, 1701– 1708 Wang, C. Y., Mayo, M. W., Korneluk, R. G., Goeddel, D. V., and Baldwin, A. S., Jr. (1998) NF-kappaB antiapoptosis: induction of TRAF1 and TRAF2 and c-IAP1 and c-IAP2 to suppress caspase8 activation. Science 281, 1680 –1683 Feltham, R., Bettjeman, B., Budhidarmo, R., Mace, P. D., Shirley, S., Condon, S. M., Chunduru, S. K., McKinlay, M. A., Vaux, D. L., Silke, J., and Day, C. L. (2011) Smac mimetics activate the E3 ligase activity of cIAP1 protein by promoting RING domain dimerization. J. Biol. Chem. 286, 17015–17028 Nakatani, Y., Kleffmann, T., Linke, K., Condon, S. M., Hinds, M. G., and Day, C. L. (2013) Regulation of ubiquitin transfer by XIAP, a dimeric RING E3 ligase. Biochem. J. 450, 629 –638 Vandenabeele, P., and Bertrand, M. J. (2012) The role of the IAP E3 ubiquitin ligases in regulating pattern-recognition receptor signalling. Nat. Rev. Immunol. 12, 833–844 Gyrd-Hansen, M., and Meier, P. (2010) IAPs: from caspase inhibitors to modulators of NF-kappaB, inflammation and cancer. Nat. Rev. Cancer 10, 561–574 Sanna, M. G., da Silva Correia, J., Ducrey, O., Lee, J., Nomoto, K., Schrantz, N., Deveraux, Q. L., and Ulevitch, R. J. (2002) IAP suppression of apoptosis involves distinct mechanisms: the TAK1/JNK1 signaling cascade and caspase inhibition. Mol. Cell. Biol. 22, 1754 –1766 Kipnis, E., Guery, B. P., Tournoys, A., Leroy, X., Robriquet, L., Fialdes, P., Neviere, R., and Fourrier, F. (2004) Massive alveolar thrombin activation in Pseudomonas aeruginosa-induced acute lung injury. Shock 21, 444 –451 Welty-Wolf, K. E., Carraway, M. S., Ortel, T. L., and Piantadosi, C. A. (2002) Coagulation and inflammation in acute lung injury. J. Thromb. Haemost. 88, 17–25 Minami, T., Sugiyama, A., Wu, S. Q., Abid, R., Kodama, T., and Aird, W. C. (2004) Thrombin and phenotypic modulation of the endothelium. Arterioscler. Thromb. Vasc. Biol. 24, 41–53 van Nieuw Amerongen, G. P., Draijer, R., Vermeer, M. A., and van Hinsbergh, V. W. (1998) Transient and prolonged increase in endothelial permeability induced by histamine and thrombin: role of protein kinases, calcium, and RhoA. Circ. Res. 83, 1115–1123 Damgaard, R. B., Fiil, B. K., Speckmann, C., Yabal, M., zur Stadt, U., Bekker-Jensen, S., Jost, P. J., Ehl, S., Mailand, N., and Gyrd-Hansen, M. (2013) Disease-causing mutations in the XIAP BIR2 domain impair NOD2-dependent immune signalling. EMBO Mol. Med. 5, 1278 –1295 Oberoi-Khanuja, T. K., and Rajalingam, K. (2012) IAPs as E3 ligases of Rac1: shaping the move. Small GTPases 3, 131–136 Geisbrecht, E. R., and Montell, D. J. (2004) A role for Drosophila IAP1-mediated caspase inhibition in Rac-dependent cell migration. Cell 118, 111–125 Sahai, E., Garcia-Medina, R., Pouyssegur, J., and Vial, E. (2007) Smurf1 regulates tumor cell plasticity and motility through degradation of RhoA leading to localized inhibition of contractility. J. Cell Biol. 176, 35–42 Liu, J., Zhang, D., Luo, W., Yu, Y., Yu, J., Li, J., Zhang, X., Zhang, B., Chen, J., Wu, X. R., Rosas-Acosta, G., and Huang, C. (2011) X-linked inhibitor of apoptosis protein (XIAP) mediates cancer cell motility via Rho GDP dissociation inhibitor (RhoGDI)dependent regulation of the cytoskeleton. J. Biol. Chem. 286, 15630 –15640


Received for publication July 23, 2013. Accepted for publication December 16, 2013.

HORNBURGER ET AL.