j ourna l homepage: www.e lsev ie r .com/ locate /ce l l s ig
AAG8 promotes carcinogenesis by activating STAT3
F
Bing Sun a, Masahiro Kawahara b, Shogo Ehata c, Teruyuki Nagamune a,b,⁎a Department of Bioengineering, Graduate School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japanb Department of Chemistry and Biotechnology, Graduate School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japanc Department of Molecular Pathology, Graduate School of Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
O
⁎ Corresponding author at: Department of ChemistrySchool of Engineering, The University of Tokyo, 7-3-1 H8656, Japan. Tel.: +81 3 5841 7328; fax: +81 3 5841 865
Please cite this article as: B. Sun, et al., Cellul
Oa b s t r a c t
a r t i c l e i n f o
16
17
18
19
20
21
22
23
24
Article history:Received 4 February 2014Received in revised form 2 April 2014Accepted 2 April 2014Available online xxxx
Keywords:AAG8STAT3Cancer
25
26
27
ED P
R
Dysregulation of signalling pathways by changes of gene expression contributes to hallmarks of cancer. The ubiq-uitously expressed chaperone protein AAG8 (aging-associated gene 8 protein, encoded by the SIGMAR1 gene) isoften found to be overexpressed in various cancers. AAG8 is involved in ER (endoplasmic reticulum)-associateddegradation and has been intensively elaborated in neuroscience. However, its rationale in carcinogenesis hasrarely been noticed. In this study, we explored the intrinsic oncogenetic roles of AAG8 in cancer cells andfound that AAG8 promoted carcinogenesis both in vitro and in vivo. We further characterized AAG8, for thefirst time to our knowledge, as a STAT3 activator and elucidated that it alternatively activated STAT3 in additionto IL6/JAK pathway. Based on these findings and a drug screening study, we demonstrated that combined inhi-bition of AAG8 and IL6/JAK signalling synergistically limits cancer cell growth. Taken together, our findingsshed light on the fundamental evidences for identification of AAG8 as anoncoprotein and potential target for can-cer prevention, as well as highlight the importance of ER proteins in contributing to JAK/STAT signaling andcarcinogenesis.
Dysregulation of signalling pathways by changes of gene expressioncontributes to hallmarks of cancer. The ubiquitously expressed chaper-one protein AAG8 (aging-associated gene 8 protein, encoded by theSIGMAR1 gene) is often found to be overexpressed in various cancers.AAG8 is predominantly expressed at the mitochondria-associated en-doplasmic reticulum (ER) membrane (MAM) and distributes dynami-cally. It modulates both MAM-specific and plasma membrane proteinsand mitochondrial metabolism [1]. In particular, AAG8 is involved inER-associated degradation [2] and has been elaborated in neuroscience[3]. Mutations of AAG8 cause neurodegenerative diseases such as amyo-trophic lateral sclerosis [4]. However, its roles in cancer have just recent-ly been noticed. We previously discovered that AAG8 antagonistspotentially inhibit melanoma cell growth and proposed AAG8 as apromising target for melanoma therapy [5]. However, the lack of gain-or loss-of-function studies has precluded a clear understanding of therationale of AAG8 in carcinogenesis.
The STAT (signal transducer and activator of transcription) familyconsists of seven members: STAT1, STAT2, STAT3, STAT4, STAT5A,STAT5B, and STAT6. STATs are pivotal in modulating cellular functionsin response to cytokines, interferons, and various growth factors,
73
74
75
76
77
and Biotechnology, Graduateongo, Bunkyo-ku, Tokyo 113-7.amune).
ar Signalling (2014), http://d
which activate JAKs (Janus kinases), leading to key tyrosine phosphory-lation on their receptors. JAKs activation allows the binding of STATs viatheir SH2 domains to these phosphotyrosine docking sites. STATs are inturn tyrosine phosphorylated, thus allowing their dimerization and ac-tivation. STATs have been shown to be controlled by several negativeregulatory mechanisms. Notably, the SOCS (suppressor of cytokine sig-nalling) family negatively regulates STAT activation [6]. STAT3 is a well-known transcription factor that has been intensively investigated incancer and immunity [7,8]. Upon phosphorylation at Y705, activatedSTAT3 translocates into the nucleus to initiate transcription. STAT3 hy-peractivation is a feature of the majority of solid cancers. However, theregulation of STAT3 activation is not fully understood.
In this study, we explored the intrinsic roles of AAG8 in cancer cellsand found that AAG8 promoted carcinogenesis both in vitro and in vivo.We further characterized AAG8, for the first time to our knowledge, as aSTAT3 activator and demonstrated that it alternatively activated STAT3in addition to IL6/JAK pathway.
2. Materials and methods
2.1. Cell lines and reagents
DLD-1, HCT116, PANC1, AGS, MKN7, MSTO211H and B16 cell lineswere obtained from American Type Culture Collection (USA). COLO205cell line was purchased from RIKEN Cell Bank (Japan). Cell culture wasmaintained in Dulbecco's Modified Eagle's Medium (DMEM) supple-mented with 10% fetal bovine serum (FBS, Life Technologies, Carlsbad,
2 B. Sun et al. / Cellular Signalling xxx (2014) xxx–xxx
UNCO
RREC
CA, USA) in a standard incubator at 37 °Cwith 5% CO2. AAG8 antagonistsBD1047 and BD1063, and AAG8 agonist PRE084 were purchased fromSanta Cruz Biotechnology (Santa Cruz, CA, USA). Matrigel™ basementmembranematrix was from BD Bioscience (Bedford, MA, USA). Recom-binant human IL-6was fromGenzyme-Techne (Minneapolis, MN, USA).Gemcitabine was from SIGMA (St Louis, MO, USA). JSI124, YM155,Ruxolitinib, JAK Inhibitor I, and JAK Inhibitor VI were included in theSCADS Inhibitor Kits.
2.2. 3D culture
3D on-top culture of cancer cells was as previously described withsome modifications [9]. Briefly, surface of 6-well plates was coatedwith pre-thawed Matrigel (500 μl/well) with a pipette tip. For eachwell, 3 × 105 or 106 cells were resuspended in 3 ml of completeDulbecco's Modified Eagle's Medium containing 5% Matrigel and pipet-ted onto the pre-coated surface. Chemicalswere added into themediumas indicated. Cells were then cultured for the indicated days before fur-ther assays. Cells were observed and photographed under a phase con-trast microscope (OLYMPUS, Tokyo, Japan).
2.3. Establishing stable cell lines
For AAG8 overexpression, the SIGMAR1 gene was cloned from thecDNA of DLD-1 cells with the forward primer 5′-ACCCAAGCTGGCTAGAATGCAGTGGGCCGTG-3′ and reverse primer 5′-GTGGATCCGAGCTCGTCAAGGGTCCTGGCCAAAG-3′ and subcloned into pcDNA3.1 vector(Life Technologies) between NheI and KpnI sites. DLD-1 and AGS cellswere transfected with either the empty vector or the plasmid express-ingAAG8 using Lipofectamine LTXwith Plus reagent (Life Technologies)according to the manufacturer's instructions. 48 h after transfection,cells were selected by 700 μg/ml G418 (WAKO, Wako, Japan). Forgene knockdown, we employed the RNA polymerase II promoter U6in pLKO.1 vector to express shRNA targeting the 5′-CCTCAACCCAGCAGCAATTTG-3′ sequence of SIGMAR1 gene. Lentivirus incorporated withshRNA was generated in HEK293T cells by combining packing plasmidpCMV-dR8.91, envelope plasmid VSV-G, (gifts from Dr. Kenneth Rock,University of Massachusetts Medical School, Worcester, MA) and thepLKO.1 plasmids. DLD-1 cells were infected with the lentivirus and se-lected by 3 μg/ml puromycin (SIGMA). HCT116 and AGS cells weretransfected directly with the shRNA plasmids using Lipofectamine LTXwith Plus reagent (Invitrogen) according to the manufacturer'sprotocol. 48 h after transfection, cells were selected by 1 μg/ml puromy-cin (SIGMA). A scramble shRNA plasmid (kindly provided by Dr. DavidSabatini, Addgene plasmid 1864) was used as the control [10].
2.4. Transient API4 knockdown
DLD-1 cells were transfected with the pLKO.1 shRNAs targeting the5′-CGTCCGGTTGCGCTTTCCTTT-3′ sequence (shAPI4-1) and the 5′-CCGCATCTCTACATTCAAGAA-3′ sequence (shAPI4-2) of BIRC5 (API4)gene using Lipofectamine LTX with Plus reagent (Invitrogen) accordingto the manufacture's protocol. A scramble shRNA plasmid was used asthe control. Seventy-two hours after transfection, cells were treated asindicated.
2.5. Growth assay and apoptosis assay
For the growth assay, dead cells were stained with trypan blue andthe total cell number was evaluated with Countess™ (Life Technolo-gies). For the apoptosis assay, cells were treated with chemicals as indi-cated in 3D Matrigel culture and then stained with 1 nM ethidiumbromide (EtBr) for 5 min. The stained DNA were observed andphotographed under a fluorescence microscope (OLYMPUS).
Please cite this article as: B. Sun, et al., Cellular Signalling (2014), http://d
ED P
RO
OF
2.6. Xenografts
Animal experimentswere performed in accordancewith the policiesof the Animal Ethics Committee of the University of Tokyo. FemaleBALB/c nu/nu mice (4 weeks of age) were purchased from the SankyoLabo Service Corporation. Three million cells with Matrigel in 100 μlwere injected into the flank of eachmouse subcutaneously, and tumorswere measured as described previously [11].
2.7. Western blot
Cells were lysed with Laemmli buffer, and each lysate sample wasloaded into two adjacent lanes, if indicated, of a 10% polyacrylamidegel forminimizing loading differences. For PKM2detection, cytoplasmicandnuclear fractionswere preparedwithNE-PERNuclear and Cytoplas-mic Extraction Reagents (Thermo Scientific, Rockford, RL, USA) accord-ing to the manual's instructions. Proteins were separated at 30 mA andtransferred onto PVDF membranes (Millipore, Darmstadt, Germany).Membranes were blocked for 1 h at room temperature using 5% skimmilk or 5% BSA (for phosphorylation detection) in TBS-Tween (TBS-T).Western blot analysis was performed according to the antibody manu-facturer's specifications. The membranes were incubated with primaryantibodies overnight in either 5% BSA or 5% skim milk in TBS-T at 4 °C.The membranes were washed thrice in TBS-T. The appropriate HRP-conjugated secondary antibody was added into 5% skim milk in TBS-T,followed by three washes in TBS-T. The membranes were developedusing a Luminata Crescendo Western HRP substrate (Millipore).
Antibodies used in this work are as follows: pSTAT3 (#9145) an-tibodies were from Cell Signalling Technology (Danvers, MA, USA).STAT3 (sc-482) antibodies were from Santa Cruz Biotechnology.AAG8 (HPA018002) and GAPDH (G9295) antibodies were fromSIGMA. API4 (NB500-201H) and LMNB (NBP1-19804) antibodieswere from NOVUS (Littleton, CO, USA). PKM2 (ab150377) antibodywas from abcam (Cambridge, MA, USA). The secondary HRP-conjugated anti-rabbit IgG antibody (G21234) was from LifeTechnologies.
2.8. Statistical analysis
All statistical analyseswere performedusingOrigin 8 Software. Errorbars indicate standard errors of themean (S.E.M.). Time courses or dosedependence was analyzed by two-tailed unpaired t-test or one-wayANOVA followed by appropriate post hoc test.
3. Results
3.1. Oncogenetic AAG8
We firstly observed that specific AAG8 antagonist BD1047 in-duced growth-suppressive phenotype of colorectal COLO205 cancercells, pancreatic PANC1 cancer cells, and gastric AGS cancer cells in3D Matrigel culture (Fig. 1A and Fig. S1A, B). Moreover, BD1047 po-tently suppressed mesothelioma MSTO211H cell growth in 3D cul-ture; in contrast, AAG8 agonist PRE084 failed to suppress growthand tube formation, although a different phenotype was observed(Fig. S1C). In addition, BD1047 induced apoptosis of 3D-culturedCOLO205 cells (Fig. S2A). BD1047 also dose-dependently suppressedthe growth of colorectal COLO205 and DLD-1 cancer cells, as well asgastric MKN7 cancer cells (Fig. S2B–D).
Gain- and loss-of-function approaches were employed for furtherconfirmation. AAG8 overexpression promoted proliferation of bothDLD-1 and gastric AGS cancer cells (Fig. 1B, left and Fig. S3). In linewith this, AAG8 knockdown with short hairpin RNA (shRNA) delayedDLD-1 cell proliferation (Fig. 1B, right) and suppressed AGS cell growthin 3D culture (Fig. S4A, B), which closely mimicked the phenotypechanges observed with AAG8 antagonist (Fig. S1A). Interestingly,
Fig. 1. Oncogenetic effects of AAG8. (A) Phase contrast images showing COLO205 cells cultured in 3DMatrigel and treated with or without 100 μM BD1047 (AAG8 antagonist) for 48 h.(B) Immunoblot of AAG8 and proliferation assay of DLD-1 cells with stable overexpression (left) or AAG8 knockdown (right). Total lysates from control and stable knockdown cell lineswere immunoblotted with AAG8 antibody; GAPDH served as the loading control. Cell numberwas counted every 3 days. Overexpression: initial cell number= 5 × 103. Knockdown: ini-tial cell number = 5 × 104. n = 3. Error bars (s.e.m.) are indicated. *p b 0.05, ***p b 0.001 (two-tailed unpaired t-test). (C) Three million DLD-1 cells were injected subcutaneously intoathymic nude mice. Representative pictures of mice are shown in left panel and quantitative measurements are shown in right. n = 7 (shCtrl), n = 8 (shAAG8). Error bars (s.e.m) areindicated. **p b 0.01 (one way repeated ANOVA followed by Tukey's test).
3B. Sun et al. / Cellular Signalling xxx (2014) xxx–xxx
although AAG8 knockdown resulted in few to no alternations of themorphogenesis of colorectal HCT116 cancer cells in 3D culture, it in-creased their sensitivity to gemcitabine, a clinical cancer drug(Fig. S4C, D). In agreement with the data in vitro, AAG8 knockdown
Please cite this article as: B. Sun, et al., Cellular Signalling (2014), http://d
slowed xenograft tumor formation of DLD-1 cells in vivo (Fig. 1C).These results collectively illustrate the tumor-promoting roles ofAAG8, implying that AAG8 serves as an oncoprotein, and strongly indi-cate AAG8 as a potential target for tumor chemotherapy.
4 B. Sun et al. / Cellular Signalling xxx (2014) xxx–xxx
3.2. Identification of AAG8 as a STAT3 activator
We unanticipatedly discovered STAT3 inactivation in both PANC1(Fig. 2A) and AGS cells treated with BD1047 in a dose-dependent man-ner (Fig. 2B), which might explain the underlying molecular mecha-nisms of AAG8 in promoting carcinogenesis. To confirm this, a time-dependent assay revealed that STAT3 activity began to decrease 3 hafter BD1047 treatment and was largely suppressed after 6 h inmouse melanoma B16 cells (Fig. 2C). We supposed that AAG8 may actas a STAT3 activator to enhance cancer cell proliferation. Supportingthis hypothesis, we found that STAT3 Y705 phosphorylation level wasincreased by ectopic AAG8 expression in DLD-1 cells (Fig. 3A, left). Con-sistently, AAG8 knockdown decreased STAT3 activation in both DLD-1cells (Fig. 3A, right), as well as in AGS cells (Fig. 3B). These findings indi-cate AAG8 as an upstream STAT3 activator.
T
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
ORREC
3.3. Dual STAT3 activation by AAG8 and JAK signalling
We next performed a SCADS (screening committee of anticancerdrugs) screening using DLD-1 cells stably expressing both AAG8 and aluciferase STAT3 reporter, such that STAT3 activity could be monitoredafter drug treatment. Among 364 chemicals with 232 targets, an API4(apoptosis inhibitor 4) inhibitor YM155, was found to dramatically(fold change N 20) decrease STAT3 activity in these cells (data notshown). This is consistent with the published data that YM155 reducedSTAT3 phosphorylation in PANC1 cells [12] and suggests that YM155might block STAT3 activation either dependent on or independent ofAAG8-related signalling. To further disentangle this event, DLD-1 cellswith stable AAG8 knockdown were treated with or without YM155for 12 h, followed by IL6 stimulation. As a result, IL6-induced robustSTAT3 activation was largely abolished by YM155 treatment (Fig. 4A),confirming the STAT3 inhibitory effect of YM155 in SCADS screeningand suggesting that YM155 disturbs the signalling activities, which areindispensable for IL6-induced STAT3 activation. In contrast, AAG8knockdown contributes no change to IL6-induced STAT3 phosphoryla-tion level (Fig. 4A), meaning that AAG8 activates STAT3 beyond IL6-dependency. Substantiating this conjecture, specific AAG8 antagonistsBD1047 or its analog BD1063 failed to decrease IL6-induced STAT3 acti-vation in DLD-1 cells (Fig. 4B). Surprisingly, AAG8 knockdown furtherdiminished the remaining pSTAT3 in YM155-treated cells (Fig. 4A).Based on the above findings, we hypothesized that AAG8 knockdowncould also decrease STAT3 activity in cells with inhibition of IL6/JAKpathway. As expected, similar results to Fig. 4A were obtained withtwo JAK inhibitors JAK inhibitor I and Ruxolitinib (Fig. 4C).
To analyze the temporal regulation of STAT3 by YM155 in AAG8-knockdown cells, we decreased the time of YM155 treatment to 6 h orincreased it to 18 h (Fig. 5A). YM155 treatment for 6 h hardly affectedAPI4; however, AAG8 knockdown led to decreased STAT3 activity in
UNC
Fig. 2. STAT3 inactivation by AAG8 antagonism. (A) Immunoblot of pSTAT3 and STAT3 in PANCSTAT3 in AGS cells in 2D culture treatedwith BD1047 of indicated concentrations for 24 h. (C) Imdifferent periods of time (0, 1, 3, 6, 9 and 12 h). Total lysates from control and treated cells we
Please cite this article as: B. Sun, et al., Cellular Signalling (2014), http://d
ED P
RO
OF
YM155-treated cells. In contrast, YM155 treatment for 18 h led to de-creased API4 expression, dramatically lowered STAT3 activity, andeven reduced total STAT3 level. Notably, AAG8 knockdown furtherabolished STAT3 activation and reduced API4 in YM155-treated cells.These data exclude the possibility that YM155 inhibits STAT3 activationdependent on AAG8-related signalings. These results together give riseto the conclusion that AAG8 alternatively activates STAT3 in addition toIL6/JAK pathway.
API4 is an anti-apoptotic protein, and its transcription can be con-currently modulated by several transcription factors such as STAT3[13], β-catenin, and YAP1 [14]. We could not establish the stable API4knockdown DLD-1 cell line, perhaps due to its cytotoxicity, as reportedelsewhere [15]. To investigate whether API4 is required for IL6-inducedSTAT3 activation, we transiently knocked down API4 in DLD-1 cellswith two specific shRNAs (Fig. 5B). To our surprise, API4 depletionwas dispensable for IL6-induced STAT3 phosphorylation in these cells(Fig. 5B). Although YM155 has been principally regarded as an API4 in-hibitor, its specificity remains uncertain. Since YM155 dramatically de-creased IL6-induced STAT3 activation while API4 knockdown did not,we supposed that YM155 might employ other inhibition mechanismsfor this inactivation. There are at least two pieces of evidencesupporting this argument. Firstly, in Fig. 5A, while YM155 treatmentfor 6 h did not decrease API4 expression, it had already cooperatedwith AAG8 knockdown to synergistically decrease STAT3 activation.Secondly, YM155 directly targets ILF3 (interleukin enhancer-bindingfactor 3), a transcription factor, to suppress API4 expression [16]. Con-clusively, some proteins besides API4 could be suppressed by ILF3-dependent YM155 treatment. In summary, YM155 appears to be a po-tent STAT3 inhibitor independent of its suppression on API4.
Furthermore, although PKM2 (pyruvate kinase M2)was reported totranslocate into the nucleus and function as a direct kinase for STAT3phosphorylation at Y705 [17], we could not detect the changes in bothexpression level and cellular distribution of PKM2 upon AAG8 overex-pression or knockdown (Fig. 6), suggesting that PKM2 might not be in-volved in AAG8-induced STAT3 activation. Taken together, our dataidentified AAG8 as an alternative STAT3 activator in addition to JAKsand PKM2 kinases.
3.4. Combined inhibition of AAG8 and JAK signalling
We then supposed that combining YM155 and AAG8 knockdowncould synergistically suppress STAT3 activation and cancer cell growth.Accordingly, we observed that YM155 treatment significantly limitedDLD-1 cell growth in 3D culture, which was enhanced by AAG8 knock-down (Fig. 7A), suggesting the synergistic antitumor effects by com-bined inhibition of these two proteins. Similarly, AAG8 knockdownsignificantly slowed the proliferation of DLD-1 cells treated with theJAK inhibitor JSI-124 (Fig. 7B) and JAK3 Inhibitor VI (Fig. 7C).
1 cells in 3D culture treated with 100 μM BD1047 for 48 h. (B) Immunoblot of pSTAT3 andmunoblot of pSTAT3 and STAT3 in B16 cells in 2D culture treatedwith 100 μMBD1047 forre immunoblotted with pSTAT3 antibody; STAT3 served as the loading control.
Fig. 3. STAT3 activation byAAG8. (A) Immunoblot of pSTAT3 and STAT3 inDLD-1 cells with stable AAG8 overexpression (left panel) or stable AAG8 knockdown (right panel). Mean valuesof pSTAT3 versus STAT3 levels were labeled with control cells as standard. (B) Immunoblot of pSTAT3 and STAT3 in stable AAG8-knockdown AGS cells. Mean values of pSTAT3 versusSTAT3 levels were labeled with control cells as standard. Total lysates from control and treated cells were immunoblotted with pSTAT3 antibody; STAT3 served as the loading control.
5B. Sun et al. / Cellular Signalling xxx (2014) xxx–xxx
4. Discussion
In this study, we show that AAG8, a chaperon protein, promotes car-cinogenesis by activating STAT3. Inhibition of AAG8 by antagonists orshRNA efficiently suppressed cancer cell growth in vitro and tumor for-mation in vivo. Furthermore, by drug screening analysis, we pinpointAAG8 as an alternative STAT3 activator in addition to IL6/JAK signalling.This study elucidates, for the first time, the critical roles of AAG8 in reg-ulating JAK/STAT3 signalling pathway, although previous studies dem-onstrated that AAG8 associated with several ion channels and/orreceptors to regulate cellular ion signalling [3,18].
UNCO
RRECT
A
B
C
shAAG8IL6
_ _ +
pSTAT3
STAT3
shCtrl shAAG8IL6
_ _ +
pSTAT3
STAT3
shCtr
shCtrl shCtrl
Fig. 4. STAT3 activation byAAG8beyond JAK signalling. (A)AAG8-knockdownDLD-1 cellswerefor 1 h. Total lysates from control and stable knockdown cell lines were immunoblottedwith pSin DLD-1 cells treated with BD1047 or BD1063 for 12 h, followed by PBS or 10 ng/ml IL6 treatmloading control. (C) Upper: AAG8-knockdown DLD-1 cells were treatedwith orwithout 1 μMRuwere non-adjacent in the gel are indicated by a vertical black line. Lower: AAG8-knockdown DL10 ng/ml IL6 treatment for 1 h. Total lysates from control and stable knockdown cells were impSTAT3 versus STAT3 levels in Ruxolitinib- or JAK inhibitor I-treated cells were labeled with sh
Please cite this article as: B. Sun, et al., Cellular Signalling (2014), http://d
OO
F
The hallucinogen N,N-dimethyltryptamine (DMT) has been identi-fied as an endogenous agonist of AAG8 [19]. Exogenously, a plethoraof ligands of AAG8 have been synthesized [20,21]. Although an AAG8antagonist has been assessed for pain treatment in phase I studies[22], few have been tested for their anti-cancer property. Our evaluationof the anti-tumor effects of AAG8 antagonists suggests the novel use ofclassical neurological drugs on cancer treatment. Promisingly, somesynthesized AAG8 ligands have been reported to specifically bind toAAG8 in the nanomolar range [1]. Further efforts are needed to deter-mine whether the anti-cancer ability of AAG8 antagonists could betranslated in vivo.
ED P
R
shAAG8shAAG8+ + +
JAK inhibitor I
shCtrl shAAG8shAAG8+ + +
Ruxolitinib
l
shCtrl
treatedwith orwithout100 nMYM155 for 12h, followedbyPBSor 10 ng/ml IL6 treatmentTAT3 antibody; STAT3 served as the loading control. (B) Immunoblot of pSTAT3 and STAT3ent for 1 h. Total lysates were immunoblotted with pSTAT3 antibody; STAT3 served as thexolitinib for 20 h, followed by PBS or 10 ng/ml IL6 treatment for 1 h. Juxtaposed lanes thatD-1 cells were treated with or without 100 nM JAK inhibitor I for 24 h, followed by PBS ormunoblotted with pSTAT3 antibody; STAT3 served as the loading control. Mean values ofCtrl cells as the standard.
Fig. 5. STAT3 regulation by YM155 and API4 knockdown. (A) Temporal STAT3 regulation by YM155. Immunoblot of pSTAT3, STAT3 andAPI4 inAAG8-knockdownDLD-1 cells treatedwithorwithout YM155 for the indicated time, followed by PBS or 10 ng/ml IL6 treatment for 1 h. Total lysateswere immunoblottedwith pSTAT3, STAT3 and API4 antibodies; GAPDH served asthe loading control. (B) Effect of API4 knockdown on IL6-induced STAT3 activation. Immunoblots of pSTAT3, STAT3 and API4 in transient API4-knockdown DLD-1 cells treatedwith PBS or10 ng/ml IL6 treatment for 1 h. Total lysates were immunoblotted with pSTAT3, STAT3 and API4 antibodies; STAT3 served as the loading control.
6 B. Sun et al. / Cellular Signalling xxx (2014) xxx–xxx
RE
At present, the exact mechanisms by which AAG8 activates STAT3 isuncertain. STAT3 phosphorylation of Y705 is concurrently and tightlycontrolled by multiple kinases and protein tyrosine phosphatases,which are a large and structurally diverse family of enzymes that
UNCO
R 320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
Fig. 6.Dispensable effects of AAG8 on PKM2. (A) Immunoblot of PKM2 in DLD-1 cells witheitherAAG8overexpression (left) or knockdown (right). Total lysateswere immunoblottedwith AAG8 antibody; GAPDH served as the loading control. (B) Immunoblot of PKM2 oftotal, cytoplasmic, and nuclear fractions in AAG8-knockdownHCT116 cells. Lysates of indi-cated fractionswere immunoblottedwith PKM2antibody. LMNB served as the loading con-trol for the nuclear fraction. GAPDH served as the loading control for total and cytoplasmicfractions.
Please cite this article as: B. Sun, et al., Cellular Signalling (2014), http://d
catalyze the dephosphorylation. We demonstrated that STAT3 kinasesare dispensable in AAG8-induced STAT3 activation. However, there isa possibility that dephosphorylation might account for AAG8-relatedregulation. For instance, PTPMeg2 is a physiologic STAT3 phosphatasethat can directly dephosphorylate STAT3 at the Tyr705 residue [23].This possibility merits future detailed evaluation.
Despite great efforts focusing on STAT3 for cancer therapy, few effi-cient strategies have been developed [24]. Single usage of chemicaldrugs targeting upstream kinases of STAT3, such as JAK2 inhibitors,often results in drug insensitivity due to acquired resistance [25]. Inthe anticancer drug screening, we identified YM155 as a potent STAT3suppressant, suggesting that YM155 blocks the signalling activitieswhich are required for IL6-induced STAT3 activation. We further illus-trated that combined inhibition of AAG8 and IL6/JAK signalling syner-gistically limits cancer cell growth. As single use of both JAK inhibitorsand API4 inhibitor YM155 has clinical limitations [25,26], our drug com-bination strategy provides a promising therapeutic approach for in-creasing the antitumor efficacy and decreasing drug resistance.
5. Conclusions
The present study characterized AAG8 as an oncoprotein in multipletypes of cancers through investigating its cancer-promoting effects andthe underlying mechanisms. We uncovered the molecular clues thatAAG8 is an alternative upstream STAT3 activator in addition to IL6/JAKsignalling pathway. Tandem inhibition of AAG8 and JAK signalling syn-ergistically suppresses cancer cell growth. Taken together, our findings
Fig. 7. Tandem AAG8–JAK inhibition. (A) Phase contrast images showing acinar morphol-ogy of the indicatedDLD-1 cells in 3D culture on day 5 after treatmentwith orwithout 100nM YM155. (B) Growth assay of AAG8-knockdown DLD-1 cells treated with 1 μM JSI-124(JAK inhibitor) for 24 h. Initial cell number = 6 × 105. n = 3. Error bars (s.e.m.) are indi-cated. ***p b 0.001 (one way ANOVA followed by Tukey's test). (C) Phase contrast imagesshowing acinar morphology of the indicated DLD-1 cells in 3D culture on day 10 aftertreatment with or without 100 nM JAK3 inhibitor VI.
7B. Sun et al. / Cellular Signalling xxx (2014) xxx–xxx
shed light on the fundamental evidences for identification of AAG8 as apotential target for cancer prevention and highlight the importance ofER chaperon proteins in contributing to JAK/STAT signalling andcarcinogenesis.
Please cite this article as: B. Sun, et al., Cellular Signalling (2014), http://d
ED P
RO
OF
Acknowledgments
We thank Dr. Daizo Koinuma and Dr. Kohei Miyazono (The Univer-sity of Tokyo) for their helpful suggestions and discussions; and Screen-ing Committee of Anticancer Drugs supported by Grant-in-Aid forScientific Research on Innovative Areas, Scientific Support Programsfor Cancer Research, from The Ministry of Education, Culture, Sports,Science and Technology, Japan for the generous gift of The SCADS Inhib-itor Kits I-IV. B. S.was supported byChina Scholarship Council (CSC) andCenter for Medical System Innovation (CMSI) of The University ofTokyo.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.cellsig.2014.04.001.
References
[1] K.S.C. Marriott, M. Prasad, V. Thapliyal, H.S. Bose, J. Pharmacol. Exp. Ther. 343 (2012)578–586.
[2] T. Hayashi, E. Hayashi, M. Fujimoto, H. Sprong, T.P. Su, J. Biol. Chem. 287 (2012)43156–43169.
[3] S. Kourrich, T. Hayashi, J.Y. Chuang, S.Y. Tsai, T.P. Su, A. Bonci, Cell 152 (2013)236–247.
[4] A. Al-Saif, F. Al-Mohanna, S. Bohlega, Ann. Neurol. 70 (2011) 913–919.[5] B. Sun, M. Kawahara, T. Nagamune, Cancer Med. (2014).[6] X.X.O. Yang, H.Y. Zhang, B.S. Kim, X.Y. Niu, J. Peng, Y.H. Chen, R. Kerketta, Y.H. Lee, S.
H. Chang, D.B. Corry, D.M. Wang, S.S. Watowich, C. Dong, Nat. Immunol. 14 (2013)(732 − +).
[7] M. Hatziapostolou, C. Polytarchou, E. Aggelidou, A. Drakaki, G.A. Poultsides, S.A.Jaeger, H. Ogata, M. Karin, K. Struhl, M. Hadzopoulou-Cladaras, D. Iliopoulos, Cell147 (2011) 1233–1247.
[8] J.W. Shui, A. Larange, G. Kim, J.L. Vela, S. Zahner, H. Cheroutre, M. Kronenberg, Na-ture 488 (2012) (222 − +).
[9] G.Y. Lee, P.A. Kenny, E.H. Lee, M.J. Bissell, Nat. Methods 4 (2007) 359–365.[10] D.D. Sarbassov, D.A. Guertin, S.M. Ali, D.M. Sabatini, Science 307 (2005) 1098–1101.[11] Y. Shirai, S. Ehata, M. Yashiro, K. Yanagihara, K. Hirakawa, K. Miyazono, Am. J. Pathol.
Ryu, J.L. Lee, J.S. Lee, T.W. Kim, PLoS ONE 7 (2012).[13] N. Sen, X.B. Che, J. Rajamani, L. Zerboni, P. Sung, J. Ptacek, A.M. Arvin, Proc. Natl.
Acad. Sci. U. S. A. 109 (2012) 600–605.[14] J. Rosenbluh, D. Nijhawan, A.G. Cox, X.N. Li, J.T. Neal, E.J. Schafer, T.I. Zack, X.X.Wang,
A. Tsherniak, A.C. Schinzel, D.D. Shao, S.E. Schumacher, B.A. Weir, F. Vazquez, G.S.Cowley, D.E. Root, J.P. Mesirov, R. Beroukhim, C.J. Kuo, W. Goessling, W.C. Hahn,Cell 151 (2012) 1457–1473.
[15] A.V. Sarthy, S.E. Morgan-Lappe, D. Zakula, L. Vernetti, M. Schurdak, J.C.L. Packer, M.G.Anderson, S. Shirasawa, T. Sasazuki, S.W. Fesik, Mol. Cancer Ther. 6 (2007) 269–276.
[16] N. Nakamura, T. Yamauchi, M. Hiramoto, M. Yuri, M. Naito, M. Takeuchi, K.Yamanaka, A. Kita, T. Nakahara, I. Kinoyama, A. Matsuhisa, N. Kaneko, H. Koutoku,M. Sasamata, H. Yokota, S. Kawabata, K. Furuichi, Mol. Cell. Proteomics 11 (2012).
[17] X.L. Gao, H.Z. Wang, J.J. Yang, X.W. Liu, Z.R. Liu, Mol. Cell 45 (2012) 598–609.[18] D. Balasuriya, A.P. Stewart, D. Crottes, F. Borgese, O. Soriani, J.M. Edwardson, J. Biol.
Chem. 287 (2012) 37021–37029.[19] D. Fontanilla, M. Johannessen, A.R. Hajipour, N.V. Cozzi, M.B. Jackson, A.E. Ruoho, Sci-
ence 323 (2009) 934–937.[20] L.X. Wang, J.A. Eldred, P. Sidaway, J. Sanderson, A.J.O. Smith, R.P. Bowater, J.R.
Reddan, I.M. Wormstone, Mech. Ageing Dev. 133 (2012) 665–674.[21] A. Hyrskyluoto, I. Pulli, K. Tornqvist, T.H. Ho, L. Korhonen, D. Lindholm, Cell Death
Dis. 4 (2013).[22] M. Abadias, M. Escriche, A. Vaque,M. Sust, G. Encina, Br. J. Clin. Pharmacol. 75 (2013)
Curran, F.R. Khuri, X.M. Deng, Mol. Cancer Ther. 12 (2013) 2200–2212.[24] M. Sen, S.M. Thomas, S. Kim, J.I. Yeh, R.L. Ferris, J.T. Johnson, U. Duvvuri, J. Lee, N.
Sahu, S. Joyce, M.L. Freilino, H.B. Shi, C.Y. Li, D. Ly, S. Rapireddy, J.P. Etter, P.K. Li, L.Wang, S. Chiosea, R.R. Seethala, W.E. Gooding, X.M. Chen, N. Kaminski, K. Pandit,D.E. Johnson, J.R. Grandis, Cancer Discov. 2 (2012) 694–705.
[25] P. Koppikar, N. Bhagwat, O. Kilpivaara, T. Manshouri, M. Adli, T. Hricik, F. Liu, L.M.Saunders, A. Mullally, O. Abdel-Wahab, L. Leung, A. Weinstein, S. Marubayashi, A.Goel, M. Gonen, Z. Estrov, B.L. Ebert, G. Chiosis, S.D. Nimer, B.E. Bernstein, S.Verstovsek, R.L. Levine, Nature 489 (2012) 155-U222.
[26] R.J. Kelly, A. Thomas, A. Rajan, G. Chun, A. Lopez-Chavez, E. Szabo, S. Spencer, C.A.Carter, U. Guha, S. Khozin, S. Poondru, C. Van Sant, A. Keating, S.M. Steinberg, W.Figg, G. Giaccone, Ann. Oncol. 24 (2013) 2601–2606.
![[Global DNA hipomethylation--the meaning in carcinogenesis]](https://static.documents.page/doc/80x56/635df79d88f33c6f8200f3fa/global-dna-hipomethylation-the-meaning-in-carcinogenesis.jpg)
