Ubl4A is required for insulin-induced Akt plasma membrane translocation through promotion of Arp2/3-dependent actin branching

Yu Zhaoa, Yuting Lina, Honghong Zhanga, Adriana Mañasa, Wenwen Tangb, Yuzhu Zhangc, Dianqing Wub, Anning Linc, and Jialing Xianga,1

aDepartment of Biology, Illinois Institute of Technology, Chicago, IL 60616; bDepartment of Pharmacology, Yale University School of Medicine, New Haven, CT 06520; and cBen May Department for Cancer Research, University of Chicago, Chicago, IL 60637

Edited by Melanie H. Cobb, University of Texas Southwestern Medical Center, Dallas, TX, and approved July 1, 2015 (received for review May 6, 2015) The serine-threonine kinase Akt is a key regulator of cell prolifer- Results ation and survival, glucose metabolism, cell mobility, and tumor- Ubl4A-Deficient Mice Display Increased Neonatal Mortality and a igenesis. Activation of Akt by extracellular stimuli such as insulin Defect in Liver Glycogen Synthesis. To better understand the bi- centers on the interaction of Akt with PIP3 on the plasma mem- ological functions of Ubl4A, we generated Ubl4A-deficient mice brane, where it is subsequently phosphorylated and activated (Fig. S1). Ubl4A knockout (KO) mice were viable but displayed by upstream kinases. However, it is not known how Akt increased neonatal mortality (occurring mostly within 24 h after is recruited to the plasma membrane upon stimulation. Here we A report that ubiquitin-like protein 4A (Ubl4A) plays a crucial role in birth) compared with their wild type (WT) littermates (Fig. 1 ). insulin-induced Akt plasma membrane translocation. Ubl4A knock- We noticed that some Ubl4A KO pups displayed signs of cyanosis A Inset out newborn mice have defective Akt-dependent glycogen syn- before death (Fig. 1 , ). The surviving Ubl4A KO newborns, thesis and increased neonatal mortality. Loss of Ubl4A results in however, appeared to be normal despite modestly low body weight the impairment of insulin-induced Akt translocation to the plasma (Fig. 1B). We wondered that increased mortality of Ubl4A KO membrane and activation. Akt binds actin-filaments and colocal- neonates might be the result of defective glucose homeostasis, as izes with actin-related protein 2 and 3 (Arp2/3) complex in the glycogenolysis and mobilization of glycogen storage from liver are membrane ruffles and lamellipodia. Ubl4A directly interacts with vital for the newborns to combat starvation-induced hypoglycemia Arp2/3 to accelerate actin branching and networking, allowing Akt (15). Indeed, glycogen staining analysis revealed that WT neonatal to be in close proximity to the plasma membrane for activation livers had typical abundant accumulation of glycogen ∼5 h after upon insulin stimulation. Our finding reveals a new mechanism birth (Fig. 1 C and D), which was quickly consumed 10 h after by which Akt is recruited to the plasma membrane for activation, birth (Fig. S2). In contrast, glycogen synthesis was significantly thereby providing a missing link in Akt signaling. lower in Ubl4A KO livers than in WT livers during the same period (Fig. 1 C and D). Consistently, phosphorylation of liver Ubl4A | Akt translocation | Arp2/3 complex | actin branching | insulin glycogen synthase (GS), which is an effector in the Akt-GS kinase 3β (GSK3β) pathway (16), was increased in the primary KO he- he serine-threonine kinase Akt (also known as protein kinase patocytes compared with the WT counterparts (Fig. 1E), sug- B) is a key regulator of cell proliferation and survival, glucose T gesting the activity of GS was decreased. Furthermore, the metabolism, cell mobility, and tumorigenesis, and its dysregulation β has been implicated in human diseases (1–3). Akt is activated by a inhibitory phosphorylation of GSK3 , which is the upstream ki- variety of extracellular stimuli including insulin (4). Upon stimula- nase of GS, was decreased in the Ubl4A KO hepatocyte cells (Fig. tion, Akt is recruited to the plasma membrane by phosphatidyli- 1E). Thus, the defect in liver glycogen synthesis may contribute to, nositol (3-5)-triphosphate (PIP3), which interacts directly with the at least in part, the increased neonatal mortality and growth Akt pleckstrin homology (PH) domain, and is subsequently phos- retardation in Ubl4A-deficient mice. phorylated at Thr308 and Ser473, and thus activated, by protein kinases such as phosphoinositide-dependent kinase-1 (PDK1) and Significance mammalian target of rapamycin (mTOR) complex 2 (mTORC2) – (5 8). It is, however, not known how Akt translocates to the plasma The signal-induced Akt membrane recruitment is a crucial step for membrane in response to extracellular signals such as insulin. its activation, but the underlying mechanism is incompletely un- Ubl4A (also known as Gdx in mammalian or Get5 in yeast) is derstood. We show that ubiquitin-like protein 4A (Ubl4A) is re- a small ubiquitin-like protein encoded by an X-linked house- quired for insulin-induced Akt membrane translocation through keeping (9). Although Ubl4A contains an ubiquitin-like direct promotion of actin-related protein 2 and 3 (Arp2/3)- domain at its N terminus, it lacks ubiquitination activity (10). In dependent actin branching, thereby ensuring glycogen synthesis yeast, Ubl4A homolog Get5 forms a complex with Get4 to assist for neonatal survival. As a novel Arp2/3-binding protein, Ubl4A Get3 for recruitment of nascent synthesized tail-anchored (TA) may play important roles in the translocation of other actin-bind- and their delivery to the endoplasmic reticulum (11, 12). ing molecules controlled by a similar mechanism, as well as a broad Ubl4A and other mammalian Get homologs are also involved in range of cellular functions related to the actin branching network. the recruitment and delivery of TA protein into the endoplasmic reticulum membrane (11, 13). Furthermore, Ubl4A can also Author contributions: Y. Zhao, D.W., and J.X. designed research; Y. Zhao, Y.L., H.Z., A.M., suppress tumorigenesis through regulation of signal transducer W.T., and Y. Zhang performed research; Y. Zhao, Y.L., D.W., A.L., and J.X. analyzed and activation of transcription 3 (13) and is involved in DNA data; and Y. Zhao, D.W., A.L., and J.X. wrote the paper. damage-mediated cell death (14). Here we report that Ubl4A is The authors declare no conflict of interest. required for insulin-induced Akt plasma membrane translocation This article is a PNAS Direct Submission. and activation through direct promotion of Arp2/3-dependent 1To whom correspondence should be addressed. Email: [email protected]. actin branching, thereby playing a critical role in regulation of This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. glycogen synthesis for neonatal survival. 1073/pnas.1508647112/-/DCSupplemental.

9644–9649 | PNAS | August 4, 2015 | vol. 112 | no. 31 www.pnas.org/cgi/doi/10.1073/pnas.1508647112 Downloaded by guest on September 28, 2021 by the significant reduction of phosphorylated endogenous Akt on the plasma membrane in Ubl4A KO cells, as analyzed using confocal microscope imaging (Fig. 2 E and F). Thus, loss of Ubl4A mainly affects Akt plasma membrane translocation.

Regulation of Akt Membrane Translocation by Ubl4A Depends on the Actin Network and the Arp2/3 Complex. To determine how Ubl4A regulates the translocation of Akt from the cytoplasm to the plasma membrane, a step about which very little is known, WT and KO primary fibroblasts were transfected with expression vector-encoding GFP-tagged Akt-PH domain (Akt-PH-GFP). Time-lapse microscopy revealed that upon insulin stimulation, Akt-PH-GFP proteins formed actin filament-like structures and appeared to be more readily recruited to the plasma membrane- ruffling regions in WT than in Ubl4A-deficient fibroblasts (Fig. 3A and Movies S1 and S2). Because Akt-PH-GFP seemed to be more extensively recruited to the structures similar to actin filaments and subsequently accumulated in membrane-ruffling regions (Fig. 3A), we examined the colocalization of F-actin and Akt-PH-GFP upon insulin stimulation. Confocal immunofluo- rescence microscopy revealed that Akt-PH-GFP proteins indeed

Fig. 1. Ubl4A-deficient mice display increased neonatal mortality and a defect in liver glycogen synthesis. (A) The survival rate of neonates within 24 h after birth. (Inset) Representative WT and KO littermates ∼5 h after birth. (B) The body weight of 3-wk-old mice; data shown are mean ± SEM. WT, n = 37; KO, n = 50. **P < 0.01. (C) Liver glycogen PAS staining of the neonates ∼5hafterbirth.(Scalebar,100μm.) (D) Quantitative analysis of liver PAS staining by matrix score. The average percentage of PAS positive area is scored as 0 (0%), 1 (25%), 3 (75%), or 4 (100%). Error bars present SD. n = 9. ***P < 0.001. (E) Immunoblotting analysis of phosphorylated and total GSK3β and GS of primary hepatocytes.

Loss of Ubl4A Results in Impaired Akt Plasma Membrane Translocation and Activation. The defect in glycogen synthesis in Ubl4A KO mice led us to investigate activation of the GSK3β upstream apical kinase Akt. Insulin-induced Akt phosphorylation at Ser473 and Thr308, both of which are required for Akt activation on the plasma membrane (17), was significantly decreased in Ubl4A KO primary hepatocytes and mouse embryonic fibroblasts (MEFs) compared with their WT counterparts (Fig. 2A). As expected, phosphorylation of S6K, an Akt substrate, was impaired in Ubl4A-deficient MEFs (Fig. S3A). In contrast, Akt phosphorylation at Thr450, which is involved in stabilization of newly synthesized Akt in the cytoplasm (18), was not affected (Fig. S3B). The impaired activation of Akt CELL BIOLOGY could be a result of defects either in Akt translocation from the cytoplasm to the plasma membrane or in its upstream components such as PIP3, mTOR, or PDK1 on the plasma membrane. The latter is unlikely. There was no significant difference in production of PIP3 between insulin-treated WT and KO fibroblasts (Fig. 2B C and Fig. S3 ). Consistently, the addition of exogenous PIP3 to the Fig. 2. Loss of Ubl4A results in impaired Akt plasma membrane trans- KO MEFs also failed to rescue Akt activation in response to insulin location and activation. (A) Activation of Akt in primary hepatocytes and stimulation (Fig. 2C). Furthermore, insulin-induced activation MEFs stimulated with insulin (2 μg/mL) for various times indicated. (B)Density of mTOR and its enzymatic activity were not affected by loss of quantitation of PIP3 in primary MEFs stimulated with insulin and immu- Ubl4A (Fig. S3 D and E), and translocation and activation of nostained using anti-PIP3 antibody (Fig. S3C). Error bars present SD. n = 30 PDK1 were only marginally reduced in Ubl4A KO MEFs (Fig. cells per group. (C) Activation of Akt in WT and KO MEFs treated with or μ S3F). In contrast, loss of Ubl4A, which resulted in the reduction without exogenous PIP3 (100 M) for the times indicated. (D) Immunoblots of the membrane Thr308- and Ser473-phosphorylated Akt, reduced of cellular fractionations of plasma-rich membrane (Memb), mitochondria D rich (Mito), or nucleus rich (Nuc) for endogenous phosphorylated and total total endogenous Akt proteins on the membrane (Fig. 2 ). In- Akt in WT and KO MEFs stimulated with or without insulin. (E)Cellular terestingly, translocation and activation of Akt to other major cel- localizations of endogenous pAkt (S473) in WT and KO MEFs. (Scale bar, lular compartments, such as mitochondria and nucleus (19, 20), were 10 μm.) (F) Membrane density analysis of pAkt shown in E. Error bars present not significantly affected (Fig. 2D). This notion is further supported SD. n = 65 cells per group. ***P < 0.001.

Zhao et al. PNAS | August 4, 2015 | vol. 112 | no. 31 | 9645 Downloaded by guest on September 28, 2021 Fig. 3. Ubl4A-mediated Akt activation depends on actin network and the Arp2/3 complex in membrane ruffles and lamellipodia. (A) Snapshots of time-lapse movies show that Akt-PH-GFP formed actin filament-like structures and translocated to the plasma membrane upon insulin stimulation. (Scale bar, 10 μm.) (B) Colocalization between transfected Akt-PH-GFP and actin. The images show a portion of membrane region of individual cells. (Inset) Line scan of fluorescence intensity. (Scale bar, 10 μm.) (C) Colocalization among endogenous pAkt, actin, and Arp2/3 in WT and KO MEFs treated with insulin. (Inset) Line scan of fluorescence intensity. (Scale bar, 10 μm.) (D) Immunoprecipitation with anti-Akt antibody in combination with immunoblotting with anti-Arp3 and anti-actin antibodies. Input, one tenth of celllysates. (E) Cosedimentation (Co-sed) assay of F-actin filaments with purified GST-Akt. (F and G) Activation of Akt in WT MEFs treated with or without actin polymerization inhibitor LatrB (2.5 μM) (F), or transfected with Arp3 siRNA (siArp3) (G). Validation of Arp3 knockdown was shown in Fig. S4B.

colocalized with F-actin filaments in WT, but not in the Ubl4A- F-actin inhibitor Latrunculin B (LatrB) (23), an Arp2/3 inhibitor deficient fibroblasts (Fig. 3B). Consistently, phosphorylated en- (CK666) (24), or Arp3 siRNA abrogated insulin-induced Akt dogenous Akt was also colocalized with actin in the plasma activation (Fig. 3 F and G and Fig. S4). These results indicate membrane, especially in membrane ruffles and lamellipodia in that actin and the Arp2/3 complex are critical for Ubl4A-mediated WT fibroblasts (Fig. 3C). Interestingly, phosphorylated endoge- Akt translocation to the plasma membrane for activation. nous Akt was well colocalized with actin-related protein complex Arp2/3, which is essential for actin branching underneath the Ubl4A Directly Interacts with Arp2/3 and Thereby Accelerates Arp2/3- plasma membrane (21, 22) in the same regions (Fig. 3C). The Dependent Actin Nucleation. The observations that both actin and colocalization of phosphorylated endogenous Akt with actin and Arp2/3 are critical for Ubl4A-mediated Akt activation led us to Arp2/3ontheplasmamembranewas significantly reduced in test whether Ubl4A regulated Arp2/3-dependent actin branching, Ubl4A-deficientMEFs(Fig.3C). In addition, Akt interacted thereby bringing Akt in close proximity to the plasma membrane only with actin, especially F-actin filaments, but not Arp2/3, as for activation. First, we analyzed the cellular localization of determined by coimmunoprecipitation and cosedimentation Ubl4A and Arp2/3 proteins. Confocal immunofluorescence (Fig. 3 D and E). Furthermore, the Ubl4A-dependent colocali- microscopy revealed that Ubl4A indeed colocalized with Arp2/3 zation of Akt with actin and Arp2/3 upon stimulation appeared in membrane ruffles and lamellipodia (Fig. 4A). To further to involve in Akt activation, as treatment of primary MEFs with determine whether there was a physical interaction between

9646 | www.pnas.org/cgi/doi/10.1073/pnas.1508647112 Zhao et al. Downloaded by guest on September 28, 2021 Fig. 4. Ubl4A directly interacts with Arp2/3 and promotes Arp2/3-dependent actin branching nucleation. (A) Colocalization between endogenous Ubl4A and Arp2/3 in WT fibroblasts upon insulin stimulation. (Scale bar, 10 μm.) (Lower) Enlarged boxed areas from regions I and II. (Inset) Quantitation of multiple-line (region I, error bars present SD, nine line scans) or single-line (region II) scans. (B) GSD-3D imaging of immunostained endogenous Ubl4A (red) and Arp2/3 (green). (Scale bar, 0.1 μm.) (C) A zoom-in view of a cross-section of individual molecule of B shows 3D direct interaction between Ubl4A and Arp2/3. (Scale bar, 10 nm.) (D) Pulldown assay using purified His-Ubl4A, actin, and Arp2/3 proteins. (E) Coimmunoprecipitation assay for Ubl4A and Arp2/3. (F) Actin poly- merization assay (Bottom) and actin branching imaged by TIRF microscopy using anti-stain 488-phalloidin-labeled actin (Top). (Scale bar, 5 μm.) (G) Pulldown assay for GST-VCA and immunoblot with anti-Arp3 antibody.

Ubl4A and the Arp2/3 complex, the 3D localizations of indi- protein). VCA is known to be essential for Arp2/3-dependent vidual Ubl4A and Arp2/3 protein molecules were visualized using actin nucleation (25–27) (Fig. 4F, b, Actin+Arp2/3+VCA). Im- CELL BIOLOGY a super resolution microscope based on ground state depletion portantly, Ubl4A was able to further promote Arp2/3-VCA- (GSD) (Fig. 4 B and C); the results demonstrate that Ubl4A dependent actin branching (Fig. 4F, c,Actin+Ubl4A+Arp2/3+ interacts directly with Arp2/3. Consistently, pulldown assay using VCA). However, Ubl4A could not replace the role of VCA in recombinant His-tagged Ubl4A and purified Arp2/3 proteins also activation of Arp2/3 (Fig. 4F, Actin+Ubl4A+Arp2/3). Instead, showed that Ubl4A specifically interacts with Arp2/3 (Fig. 4D). the interaction between Arp2/3 and VCA was further accelerated Importantly, the interaction between endogenous Ubl4A and the in the presence of Ubl4A (Fig. 4G). Taken together, this indicates Arp2/3 complex could be further augmented by insulin (Fig. 4E). that Ubl4A promotes actin branching through direct association These data suggest there is a regulatable physical interaction with the Arp2/3 complex, thereby facilitating Arp2/3 interaction between Ubl4A and Arp2/3 in response to extracellular signals. with VCA for actin nucleation upon insulin stimulation. Next, we determined whether physical interaction of Ubl4A with the Arp2/3 complex influences the actin nucleated branch- Discussion ing process. In vitro actin polymerization assay showed that ad- A key step in Akt activation is its translocation from the cyto- dition of Ubl4A alone has no detectable effect on actin assembly plasm to the plasma membrane, where it interacts with PIP3 (Fig. 4F, a, Actin+Ubl4A). The actin branch nucleation can be through the Akt PH domain and is subsequently activated by initiated by Arp2/3 in the presence of VCA (verprolin, cofilin, upstream protein kinases including PDK1 and mTORC2 (7, 8). acidic), a domain of WASP protein (Wiskott–Aldrich syndrome How Akt translocates from the cytosol to the plasma membrane

Zhao et al. PNAS | August 4, 2015 | vol. 112 | no. 31 | 9647 Downloaded by guest on September 28, 2021 is not known, even though it has been thought that passive dif- Glycogen Staining. Livers and hearts were harvested from newborn pups 3–10 h fusion might be involved. Our data clearly demonstrate that postbirth. Glycogen was stained using a standard Periodic acid–Schiff (PAS) Ubl4A is crucial for insulin-induced Akt activation through method in the Pathology Facility at the University of Chicago. promotion of an Arp2/3-mediated actin network underneath the Immunoblotting. Cells were lysed in Nonidet P-40 buffer (145 mM NaCl, 5 mM plasma membrane, especially in the regions of membrane ruffles MgCl2, 25 mM Hepes at pH 7.4, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 0.2% and lamellipodia. Nonidet P-40), plus a mixture of protease inhibitors including 10 μg/mL The actin network (meshwork) underneath the plasma mem- leupeptin, 10 μg/mL aprotinin, 10 mM NaF, and 2 mM sodium orthovanadate. brane, especially in lamellipodia and membrane ruffles, plays a Equal amounts of proteins were separated by standard SDS/PAGE and trans- crucial role in many important cellular events (28, 29). Here we ferred to PVDF membrane. The antibodies against phospho-Akt (Ser473), show that the actin network is also important for membrane phospho-Akt (Thr308), phospho-Akt (Thr450), Akt, phospho-p70 S6 Kinase β β translocation of Akt in response to insulin. This conclusion is sup- (Thr389), p70 S6 Kinase, phospho-GSK3 , GSK3 , phospho-Glycogen Syn- thase (pGS, Ser641), Arp3, phospho-mTOR (Ser2448), and mTOR were pur- ported by the following evidence. First, upon stimulation by insulin, chased from Cell Signaling. Antibodies against GS and Ubl4A were Akt translocated to the plasma membrane, clustered in lamellipodia purchased from Sigma. Representative results obtained from at least three and membrane ruffles (Figs. 2E and 3C), consistent with previous independent experiments are presented. reports (30). Second, Akt associated with actin, especially F actin filaments (Fig. 3 D and E), which is also consistent with previous Cellular Fractionation. Ten culture dishes (100 mm) of primary MEFs (

9648 | www.pnas.org/cgi/doi/10.1073/pnas.1508647112 Zhao et al. Downloaded by guest on September 28, 2021 300 μL binding buffer containing purified actin (20 μg; Cytoskeleton, Inc.) and intensities were quantitated by single-line or radial lines scan analysis by Arp2/3 complex (10 μg; Cytoskeleton, Inc.), and then the mixture was in- ImageJ. For GSD imaging, cells were postfixed with 4% paraformaldehyde cubated for 2 h at 4 °C. For the GST pulldown assay, the GST-VCA domain of for 15 min before imaging to prevent dissociation of secondary antibodies. WASP (160 nM) was incubated with Arp2/3 complex (10 nM) in polymeri- Slides were embedded with freshly prepared 20 mM β-mercaptoethylamine

zation buffer (5 mM Tris·HCl at pH 7.5, 50 mM KCl, 2 mM MgCl2, 1 mM EGTA, (Sigma) and imaged using a Leica GSD-3D microscope with a 160×/1.43 oil 0.1 mM CaCl2, 0.3 mM ATP, 0.5 mM DTT) with or without purified Ubl4A HCX PL APO objective and an Andor-DU897 BV-7849 camera. (100 μM) for 5 min at 37 °C. GST-VCA and its associated proteins were col- lected by glutathione beads. The beads were washed three times and sub- Time-Lapse Movie. Primary MEFs were seeded onto 35-mm poly-D-lysine- jected to immunoblotting analysis. coated glass bottom microwell dishes and transfected with an Akt-PH-GFP construct (Addgene plasmid 18836) for 16 h, followed by starvation and Actin Polymerization Assay and TIRF Imaging. The actin polymerization assay insulin treatment as described earlier. Time-lapse movies were recorded at was performed according to the manufacturer’s instruction (Cytoskeleton 10-s intervals for 10 min in a heated chamber at 37 °C, using a Marianas Inc.). Briefly, the reaction was initiated in an actin polymerization buffer fluorescence confocal microscope with a 488-nm solid-state laser. (5 mM Tris·HCl at pH 7.5, 50 mM KCl, 2 mM MgCl2, 1 mM EGTA, 0.1 mM CaCl2, 0.3 mM ATP, 0.5 mM DTT) containing pyrene-labeled rabbit muscle Neutrophil Chemotaxis Assay. The chemotaxis assay using a Dunn chamber actin (0.8 μM G-actin), 10 nM Arp2/3 complex, and 160 nM GST-VCA either was carried out as previously described (37), with some modifications (34, 38). μ with or without Ubl4A (100 M). Pyrene fluorescence signals were moni- Both WT and Ubl4A-deficient neutrophils were simultaneously labeled with tored using a Spectra Max GEMINI EM Spectrofluorometer (Molecular De- different tracing dyes. The experimental labeled groups were alternated in vices) at 37 °C, with excitation and emission at 350 and 405 nm, respectively. the study to completely eliminate the possibility of any influence from the For visualization of actin polymers, the reaction mixture was diluted 50-fold dye. In addition, the chemoattractant gradients were monitored by using with the polymerization buffer containing 70 nM Anti-stain 488 phalloidin free fluorescein isothiocyanate dye. Only cells under similar gradients were (Cytoskeleton). The images were captured using a TIRF microscope with analyzed, using the formula described previously (39), to calculate average × 160 /1.43 oil HCX PL APO objective. directional error, average turning angle, and motility. Time-lapse image series were acquired at 30-s intervals for 30 min. Confocal Microscopy and GSD-3D Imaging. Primary MEFs (

1. Downward J (2004) PI 3-kinase, Akt and cell survival. Semin Cell Dev Biol 15(2): 20. Bijur GN, Jope RS (2003) Rapid accumulation of Akt in mitochondria following 177–182. phosphatidylinositol 3-kinase activation. J Neurochem 87(6):1427–1435. 2. Hemmings BA, Restuccia DF (2012) PI3K-PKB/Akt pathway. Cold Spring Harb Perspect 21. Pollard TD, Borisy GG (2003) Cellular motility driven by assembly and disassembly of Biol 4(9):a011189. actin filaments. Cell 112(4):453–465. 3. Plas DR, Thompson CB (2005) Akt-dependent transformation: There is more to growth 22. Rotty JD, Wu C, Bear JE (2013) New insights into the regulation and cellular functions than just surviving. Oncogene 24(50):7435–7442. of the ARP2/3 complex. Nat Rev Mol Cell Biol 14(1):7–12. 4. Lawlor MA, Alessi DR (2001) PKB/Akt: A key mediator of cell proliferation, survival 23. Spector I, Shochet NR, Kashman Y, Groweiss A (1983) Latrunculins: Novel marine toxins that and insulin responses? J Cell Sci 114(Pt 16):2903–2910. disrupt microfilament organization in cultured cells. Science 219(4584):493–495. 5. Manning BD, Cantley LC (2007) AKT/PKB signaling: Navigating downstream. Cell 24. Hetrick B, Han MS, Helgeson LA, Nolen BJ (2013) Small molecules CK-666 and CK-869 129(7):1261–1274. inhibit actin-related protein 2/3 complex by blocking an activating conformational – 6. Franke TF (2008) Intracellular signaling by Akt: Bound to be specific. Sci Signal 1(24):pe29. change. Chem Biol 20(5):701 712. 7. Alessi DR, et al. (1997) 3-Phosphoinositide-dependent protein kinase-1 (PDK1): 25. Takenawa T, Suetsugu S (2007) The WASP-WAVE protein network: Connecting the – Structural and functional homology with the Drosophila DSTPK61 kinase. Curr Biol membrane to the cytoskeleton. Nat Rev Mol Cell Biol 8(1):37 48. 7(10):776–789. 26. Weisswange I, Newsome TP, Schleich S, Way M (2009) The rate of N-WASP exchange limits – 8. Facchinetti V, et al. (2008) The mammalian target of rapamycin complex 2 controls the extent of ARP2/3-complex-dependent actin-based motility. Nature 458(7234):87 91. folding and stability of Akt and protein kinase C. EMBO J 27(14):1932–1943. 27. Campellone KG, Welch MD (2010) A nucleator arms race: Cellular control of actin – 9. Toniolo D, Persico M, Alcalay M (1988) A “housekeeping” gene on the assembly. Nat Rev Mol Cell Biol 11(4):237 251. 28. Ridley AJ, et al. (2003) Cell migration: Integrating signals from front to back. Science encodes a protein similar to ubiquitin. Proc Natl Acad Sci USA 85(3):851–855. 302(5651):1704–1709. 10. Xu Y, Cai M, Yang Y, Huang L, Ye Y (2012) SGTA recognizes a noncanonical ubiquitin- 29. Koestler SA, et al. (2013) Arp2/3 complex is essential for actin network treadmilling as like domain in the Bag6-Ubl4A-Trc35 complex to promote endoplasmic reticulum- well as for targeting of capping protein and cofilin. Mol Biol Cell 24(18):2861–2875. associated degradation. Cell Reports 2(6):1633–1644. CELL BIOLOGY 30. Watton SJ, Downward J (1999) Akt/PKB localisation and 3′ phosphoinositide gener- 11. Mariappan M, et al. (2010) A ribosome-associating factor chaperones tail-anchored ation at sites of epithelial cell-matrix and cell-cell interaction. Curr Biol 9(8):433–436. membrane proteins. Nature 466(7310):1120–1124. 31. Xue G, Hemmings BA (2013) PKB/Akt-dependent regulation of cell motility. J Natl 12. Schuldiner M, et al. (2008) The GET complex mediates insertion of tail-anchored Cancer Inst 105(6):393–404. proteins into the ER membrane. Cell 134(4):634–645. 32. Enomoto A, et al. (2005) Akt/PKB regulates actin organization and cell motility via 13. Wang Y, et al. (2014) GdX/UBL4A specifically stabilizes the TC45/STAT3 association Girdin/APE. Dev Cell 9(3):389–402. and promotes dephosphorylation of STAT3 to repress tumorigenesis. Mol Cell 53(5): 33. Cenni V, et al. (2003) Targeting of the Akt/PKB kinase to the actin skeleton. Cell Mol – 752 765. Life Sci 60(12):2710–2720. 14. Krenciute G, et al. (2013) Nuclear BAG6-UBL4A-GET4 complex mediates DNA damage 34. Xu W, et al. (2010) Integrin-induced PIP5K1C kinase polarization regulates neutrophil – signaling and cell death. J Biol Chem 288(28):20547 20557. polarization, directionality, and in vivo infiltration. Immunity 33(3):340–350. 15. Darlington GJ (1999) Molecular mechanisms of liver development and differentiation. 35. Klaunig JE, et al. (1981) Mouse liver cell culture. I. Hepatocyte isolation. In Vitro – Curr Opin Cell Biol 11(6):678 682. 17(10):913–925. 16. Cross DA, Alessi DR, Cohen P, Andjelkovich M, Hemmings BA (1995) Inhibition of 36. Pearce LR, Sommer EM, Sakamoto K, Wullschleger S, Alessi DR (2011) Protor-1 is required glycogen synthase kinase-3 by insulin mediated by protein kinase B. Nature 378(6559): for efficient mTORC2-mediated activation of SGK1 in the kidney. Biochem J 436(1):169–179. 785–789. 37. Zicha D, Dunn G, Jones G (1997) Analyzing chemotaxis using the Dunn direct-viewing 17. Alessi DR, et al. (1996) Mechanism of activation of protein kinase B by insulin and IGF-1. chamber. Methods Mol Biol 75:449–457. EMBO J 15(23):6541–6551. 38. Zhang Y, et al. (2013) A network of interactions enables CCM3 and STK24 to co- 18. Oh WJ, et al. (2010) mTORC2 can associate with ribosomes to promote cotranslational ordinate UNC13D-driven vesicle exocytosis in neutrophils. Dev Cell 27(2):215–226. phosphorylation and stability of nascent Akt polypeptide. EMBO J 29(23):3939–3951. 39. Zhang Y, et al. (2010) Different roles of G protein subunits beta1 and beta2 in neu- 19. Xuan Nguyen TL, et al. (2006) Akt phosphorylation is essential for nuclear translocation trophil function revealed by gene expression silencing in primary mouse neutrophils. and retention in NGF-stimulated PC12 cells. Biochem Biophys Res Commun 349(2):789–798. J Biol Chem 285(32):24805–24814.

Zhao et al. PNAS | August 4, 2015 | vol. 112 | no. 31 | 9649 Downloaded by guest on September 28, 2021