C A N C E R

The kinase activity of the Ser/Thr kinase BUB1promotes TGF-b signalingShyam Nyati,1,2 Katrina Schinske-Sebolt,2 Sethuramasundaram Pitchiaya,3,4

Katerina Chekhovskiy,2 Areeb Chator,2 Nauman Chaudhry,2 Joseph Dosch,5

Marcian E. Van Dort,5 Sooryanarayana Varambally,6 Chandan Kumar-Sinha,6,7

Mukesh Kumar Nyati,2 Dipankar Ray,2 Nils G. Walter,3,4 Hongtao Yu,8,9

Brian Dale Ross,1,5 Alnawaz Rehemtulla1,2*

Transforming growth factor–b (TGF-b) signaling regulates cell proliferation and differentiation, whichcontributes to development and disease. Upon binding TGF-b, the type I receptor (TGFBRI) binds TGFBRII,leading to the activation of the transcription factors SMAD2 and SMAD3. Using an RNA interference screenof the human kinome and a live-cell reporter for TGFBR activity, we identified the kinase BUB1 (buddinguninhibited by benzimidazoles-1) as a key mediator of TGF-b signaling. BUB1 interacted with TGFBRI inthe presence of TGF-b and promoted the heterodimerization of TGFBRI and TGFBRII. Additionally, BUB1interacted with TGFBRII, suggesting the formation of a ternary complex. Knocking down BUB1 preventedthe recruitment of SMAD3 to the receptor complex, the phosphorylation of SMAD2 and SMAD3 and theirinteraction with SMAD4, SMAD-dependent transcription, and TGF-b–mediated changes in cellular pheno-type including epithelial-mesenchymal transition (EMT), migration, and invasion. Knockdown of BUB1 alsoimpaired noncanonical TGF-b signaling mediated by the kinases AKT and p38 MAPK (mitogen-activatedprotein kinase). The ability of BUB1 to promote TGF-b signaling depended on the kinase activity of BUB1.A small-molecule inhibitor of the kinase activity of BUB1 (2OH-BNPP1) and a kinase-deficient mutant ofBUB1 suppressed TGF-b signaling and formation of the ternary complex in various normal and cancer celllines. 2OH-BNPP1 administration to mice bearing lung carcinoma xenografts reduced the amount of phos-phorylated SMAD2 in tumor tissue. These findings indicated that BUB1 functions as a kinase in the TGF-bpathway in a role beyond its established function in cell cycle regulation and chromosome cohesion.

INTRODUCTION

The transforming growth factor–b (TGF-b) family of cytokines regulatesmany processes such as immune suppression, angiogenesis, wound heal-ing, and epithelial-mesenchymal transition (EMT) (1, 2). Abnormal TGF-bsignaling is linked to autoimmune and inflammatory diseases, fibrosis,tumor formation and metastasis, and various other disorders. Early intumorigenesis, the proliferation of epithelial cells retains exquisite sen-sitivity to TGF-b, wherein TGF-b elicits a tumor-suppressive response.However, transformed cells become refractory to TGF-b–mediated growthinhibition and acquire a phenotype wherein the intracellular signalingcircuitry is altered, leading to tumorigenic and metastatic effects in re-sponse to TGF-b (3). These responses are diverse, depending on signal-ing from growth factor receptors activated in parallel (4), cell density andcell cycle phase (5), and the abundance and activity of transcriptionalregulators (6).

BUB1 (budding uninhibited by benzimidazoles-1) is a serine/threoninekinase that is encoded by theBUB1 gene [40 kilo–base pairs (kbp), 25 exons]

in humans. During mitosis, BUB1 binds kinetochores and plays a key rolein establishing the mitotic spindle checkpoint and aligning chromosomes(7), in addition to its central role in ensuring fidelity during chromosomalsegregation into daughter cells (8). Three main regions have been identifiedin BUB1: a conserved N-terminal region that contains the kinetochore lo-calization domain; an intermediate, nonconserved region that is required asa scaffold for the recruitment of proteins; and aC-terminal region that containsa catalytic serine/threonine kinase domain (9). Mutations in BUB1 are asso-ciated with aneuploidy and several types of cancer.

Ligand-dependent activation of TGF-b receptors and regulation of theirsubsequent kinase activity is a complex process that can involve severalposttranslational modifications of the receptors [including autophosphoryl-ation (10), cross-phosphorylation (11), trans-phosphorylation (12), ubiqui-tylation (13), sumoylation (14), and dephosphorylation (15)], as well asinternalization of the receptor-ligand complex (16–19). With so muchcomplexity in one pathway, we set out to find potentially new regulatorsof TGF-b receptor activity.

RESULTS

Kinase-targeted high-throughput small interfering RNAscreen identifies regulators of TGF-b signalingTo identify proteins that regulate signaling by the type I TGF-b receptor(TGFBRI), we transfected small interfering RNAs (siRNAs) againsteach of the 720 known and predicted kinases into A549 lung adenocar-cinoma and MDA-231-1833 breast cancer cells that stably expressed areporter for TGFBRI kinase activity [(BTR (4)] and screened for an in-crease in bioluminescence after the addition of TGF-b ligand to the culture

1Center for Molecular Imaging, University of Michigan, Ann Arbor, MI 48109,USA. 2Department of Radiation Oncology, University of Michigan, Ann Arbor,MI 48109, USA. 3SingleMoleculeAnalysis inReal-Time (SMART)Center, Univer-sity ofMichigan, AnnArbor,MI 48109,USA. 4Department of Chemistry, Universityof Michigan, Ann Arbor, MI 48109, USA. 5Department of Radiology, University ofMichigan, Ann Arbor, MI 48109, USA. 6Department of Pathology, University ofMichigan, Ann Arbor, MI 48109, USA. 7Michigan Center for Translational Pathol-ogy, University of Michigan, Ann Arbor, MI 48109, USA. 8Howard Hughes Med-ical Institute, University of Texas Southwestern Medical Center, Dallas, TX 75390,USA. 9Department of Pharmacology, University of Texas Southwestern MedicalCenter, Dallas, TX 75390, USA.*Corresponding author. E-mail: alnawaz@med.umich.edu

R E S E A R C H A R T I C L E

www.SCIENCESIGNALING.org 6 January 2015 Vol 8 Issue 358 ra1 1

on January 6, 2015http://stke.sciencem

ag.org/D

ownloaded from

medium (fig. S1A). BTR is a reporter that exhibits increased lumines-cence upon loss of TGFBRI kinase activity; thus, positive hits from thesiRNA screen identify genes that encode proteins that promote TGFBRIkinase activity.We used a quartile-basedmethod to identify proteins that,when knocked down, inducedmore than 7.14-fold activation of the reporterin A549 cells and more than 13.91-fold in MDA-231-1833 cells (high-stringency hits, targeted error rate a = 0.0027), followed by experiment-and plate-wise analysis of the normalized fold induction (Fig. 1, A to C,and table S1) as previously described (20). Eight hits were identified inA549-BTR cells (Fig. 1A), whereas 15were recovered inMDA-231-1833–BTR cells (Fig. 1B). Two of these hits were the kinases BMPR2 (in A549-BTR cells) and RPS6KB1 (in MDA-231-1833–BTR cells), which arealready associated with the TGF-b pathway (table S2). Analysis at a lowerstringency (5.4-fold in A549 cells, 10.29-fold in MDA-231-1833 cells, tar-geted error rate a = 0.046) revealed additional hits, several of which alsohave known roles in regulating TGF-b signaling (table S2), thereby furthervalidating the screen.

A logical relations analysis of the A549-BTR and MDA-231-1833–BTR screens revealed that a large majority of the significant hits were dis-tinct for each cell line (Fig. 1C); of these, seven were common (Fig. 1C).Signaling pathway impact analysis demonstrated that signaling networkspreviously implicated in promoting the cellular response to TGF-b [includ-ing those mediated by receptor tyrosine kinases (RTKs), mitogen-activatedprotein kinase (MAPK), and phosphoinositide 3-kinase (PI3K)–AKT]wereidentified as hits in the screen (table S3 and fig. S1, B and C) (4). Targetswithin the progesterone-mediated oocytematuration and the oocytemeiosispathways, in which BUB1 is a key component, consistently exhibited highperturbation accumulation in A549-BTR and MDA-231-1833–BTR cells(fig. S1, B and C).

BUB1 promotes canonical and noncanonicalTGF-b signalingIn an effort to biochemically validate these findings, we knocked downselect hits inMRC5 (human fetal lung fibroblast), Het1A (human esoph-ageal epithelial), A549-BTR (lung adenocarcinoma), MDA-231-1833–BTR (breast cancer), MCF7 (breast cancer), and MCF10A (humanmammary epithelial) cells. BUB1 knockdown induced the most robustdecrease in TGF-b–dependent SMAD2 phosphorylation (fig. S2) andwas selected for further study. Similar to a positive control, in whichTGFBRI was silenced, knockdown of BUB1 in lung cancer (A549and NCI-H358) and breast cancer (MDA-231-1833) cell lines using siRNAsabrogated the activation of SMAD2 and SMAD3, as well as componentsof the noncanonical signaling cascade (c-JUN, p38MAPK, and AKT) ineach of the cell lines (Fig. 1D), key downstream effectors of TGF-b signal-ing (21, 22). Transfection with BUB1 siRNA caused a similar decrease inTGFBRI abundance as did TGFBRI siRNA, but only in A549 cells. Evenstill, in this cell line, the effects of BUB1 siRNAdid notmimic the effects ofTGFBRI siRNA. Thus, the effects of BUB1 knockdown on downstreammarkers of TGF-b signaling were not an artifact of off-target or indirectknockdown of TGFBRI.

BUB1 mediates TGF-b–dependent recruitmentof R-SMADs to the activated receptorBecause depletion of BUB1 reduced TGF-b–mediated SMAD2/3 phos-phorylation, we hypothesized that BUB1 may mediate the efficient re-cruitment of SMAD2/3 to the activated receptor in the presence ofligand. To test this hypothesis, A549 and human embryonic kidney(HEK) 293T cells were transfected with Flag-tagged SMAD3 and His-tagged TGFBRI in the presence of control or TGFBRI- or BUB1-targetedsiRNA and subsequently treated with TGF-b for 1 hour. Coimmuno-

precipitation analysis revealed that either TGFBRI or BUB1 knockdownimpaired SMAD3 binding to TGFBRI in the presence of ligand (Fig. 2, Aand B, and fig. S3, A to C).

Upon phosphorylation, receptor-regulated SMADs (R-SMADs, includ-ing SMADs 1, 2, 3, 5, and 8/9) translocate to the nucleus, wherein theyregulate transcription of specific target genes. We evaluated the effect ofBUB1 knockdown on the nuclear translocation of SMAD2 as a surrogatefor the activation of the TGF-b pathway by immunofluorescence microsco-py. A549 and H358 cells were transfected with control or TGFBRI- orBUB1-targeted siRNA stimulated with TGF-b for 1 hour. In control cells,SMAD2 was predominantly localized to the nucleus upon TGF-b stimula-tion; in contrast, in cells depleted of either TGFBRI or BUB1, SMAD2translocation to the nucleus was similarly impaired (fig. S3D). Together,these results suggest that BUB1plays an important role in promoting canon-ical and noncanonical TGF-b signaling.

BUB1 promotes R-SMAD/SMAD4 complex formation andtarget gene transcriptionTo extend our finding that depletion of BUB1 leads to reduced R-SMADrecruitment to the receptor complex and hence reduced phosphorylationand nuclear translocation, we investigated whether BUB1 also promotesSMAD2/3-SMAD4 complex formation using coimmunoprecipitationassays. As expected, depletion of TGFBRI in A549 cells impairedTGF-b–induced formation of a SMAD2/3-SMAD4 complex (Fig. 2Cand fig. S4). To evaluate a role for BUB1 in TGF-b–induced gene ex-pression, we used the SBE4-Luc reporter (6). In A549 cells transfectedwith control siRNA, TGF-b induced a time-dependent increase in reporteractivity, but this induction was abrogated by depletion of BUB1 (Fig. 2, Dand E). Similar observations were seen in lung (NCI-H358), breast (MDA-231-1833), and cervical (HeLa) cancer cell lines (Fig. 2, F to H).

One of the expression signatures induced by TGF-b signaling is thatwhich promotes EMT, a process in epithelial cells that occurs during devel-opment and disease progression in response to specific stimuli (23). Addi-tion of TGF-b to epithelial cells in culture induces features of EMT,including the adoption of a spindle-like mesenchymal morphology andinvasive behavior (24–26). Therefore, we investigated the role of BUB1inTGF-b–mediatedEMT.As expected, cultures ofA549orNCI-H358 cellsexhibited a transition to a mesenchymal phenotype in response to TGF-b;however, TGFBRI or BUB1 knockdown prevented this transition (fig. S5,A to E). Similar to the effect of silencing TGFBRI, knocking down BUB1significantly inhibited TGF-b–mediated induction of cell migration and in-vasion in culture (fig. S6, A to D). These results further support a role forBUB1 in mediating TGF-b signaling.

Catalytic activity is required for BUB1 to mediateTGF-b signalingTo confirm that the kinase activity of BUB1mediated TGF-b signaling, weconducted rescue experiments in A549 cells using siRNA-resistantconstructs of BUB1 (27). In cells depleted of endogenous BUB1, restora-tion with wild-type BUB1, but not a kinase-deficient BUB1 mutant, res-cued SMAD2/3 phosphorylation in a ligand-dependent manner (Fig. 3Aand fig. S7A). In agreement with this observation, reporter activity inBUB1-depleted A549 cells expressing the SBE4-Luc promoter was re-stored upon expression of wild-type, but not a kinase-deficient, BUB1(Fig. 3B), demonstrating a requirement for BUB1’s kinase activity in its rolein TGF-b signaling. The concentration of 2OH-BNPP, a catalytic BUB1inhibitor (9), negatively correlated with the phosphorylation of SMAD2and SMAD3 (fig. S7, B to D). Blocking the activity of BUB1 with 2OH-BNPP1 dose-dependently impaired the TGF-b–induced phosphorylation ofproteins that mediate canonical and noncanonical TGF-b signaling in

R E S E A R C H A R T I C L E

www.SCIENCESIGNALING.org 6 January 2015 Vol 8 Issue 358 ra1 2

on January 6, 2015http://stke.sciencem

ag.org/D

ownloaded from

0

4

8

12

16

20

0 120 240 360 480 600 720

A549

BTR

fold

chan

ge18

33 B

TR fo

ld ch

ange

BUB1

DGKQ

BMPR2BRDT

GUCY2C

EPHB4

PLK3

A

B

C

JAK2

Target number

0 120 240 360 480 600 720Target number

0

10

20

30

40

50GRK1

NYD-SP25

BLKFASKT

PI4KII RPS6KB1

KSR2

NEK6

MAP3K15

FRDA

BRD4 AATK

TSK5

BUB1

JAK2

D

PANK1CSNK1G3GRK1NME5PRKAA1BUB1PI4KIIMAPK9FASTKPRKAG1HSPB8MAPK3NEK6MAP3K15DYRK2NYD-SP25BLKDGKQEPHB6RPS6KB1KSR2

CITRPS6KA6SRPK1RAGEBMPR2NRBP2LMTK2BRD4FRDACDC2L5ALKAATKERBB3CRKLSTK4CSNK1A1PIK3CDPFKFB2BRDTGUCY2CZAKPLK3XYLBTSKSEPHB4ITPKB

0 12

A549 BTR 1833 BTRa b c a b c

JAK2

siBUB1siTGFBRI

NSS

BUB1

GAPDH

TGFBRI

pc-JUN

pP38MAPK

pAKT

AKT

1833H358A549

pSMAD3

SMAD3

pSMAD2

SMAD2

(1 hour)

Fig. 1. BUB1 mediates TGF-b ligand–dependentSMAD2/3 phosphorylation, as well as MAPKand AKT activation. (A and B) Plot of the induc-tion of a TGFBR1 reporter (BTR) in A549 cells(A) and MDA-231-1833–BTR (B) cells trans-fected with siRNAs targeting human kinases.Red or blue triangles and circles indicatehigh-stringency (triangles) or low-stringency(circles) hits either in plate- and experiment-wise analyses (red) or in experiment-wise anal-ysis only (blue). Data are representative of

three independent experiments. (C) Heat map of low-stringency hits in A549-BTR and MDA-231-1833–BTR cells from triplicate experiments. Gene namein italics indicates common hits. (D) Western blot analysis of TGF-b effector molecules in A549, NCI-H358, and MDA-231-1833 cells transfected with siRNAagainst either BUB1 (siBUB1) or TGFBRI (siTGFBRI) or a scrambled control siRNA [nonsilencing siRNA (NSS)] in the presence of TGF-b ligand for 1 hour.Blots are representative of three independent experiments.

R E S E A R C H A R T I C L E

www.SCIENCESIGNALING.org 6 January 2015 Vol 8 Issue 358 ra1 3

on January 6, 2015http://stke.sciencem

ag.org/D

ownloaded from

multiple cancer cell lines (Fig. 3, C to E) and normal cell lines (fig. S7, Eand F). Similar to the effects of a pharmacological TGFBRI inhibitor[SB431542 (28)], 2OH-BNPP impaired the nuclear translocation of

SMAD2 (fig. S7, G and H), SMAD2/3-SMAD4 complex formation (Fig. 3,F and G), and SBE4-Luc reporter activity (Fig. 3, H and I, and fig. S7, Iand J) in response to TGF-b in various cell lines.

BUB1 colocalizes and interactswith TGFBRI in response toTGF-b stimulationOn the basis of the observed role of BUB1in promoting the recruitment of SMAD3 toTGFBRI (Fig. 2, A and B, and fig. S3, Aand B) and previous evidence of a scaffold-ing function for BUB1 (7–9, 29), we inves-tigated whether BUB1 physically interactswith TGFBRI to mediate its activity usingtotal internal reflection fluorescence (TIRF)microscopy (30). TIRF images revealedthe presence ofBUB1 andTGFBRI in punc-tate structures within the plasma membraneplane (Fig. 4A). Spectral analysis detectednegligible colocalization of TGFBRI andBUB1 at 1 and 24 hours after ligand stim-ulation (fig. S8A), possibly because ofthe signal from the relatively high abun-dance of cytoplasmic BUB1. In contrast, by72 hourswith TGF-b, the proportion of co-localized TGFBRI and BUB1 signals dou-bled (Fig. 4, A to C).

We next examined whether forced over-expression of BUB1 using a Myc-taggedconstruct was sufficient for its interactionwith TGFBRI, and if this would affect thephosphorylation of SMAD2/3. Heterologousexpression of BUB1 in A549 cells resultedin the coimmunoprecipitation of BUB1 andTGFBRI regardless of TGF-b ligand (Fig.4D and fig. S8B), but this was not sufficientto initiate SMAD2/3 phosphorylation in theabsence of ligand, only modestly increasedphosphorylationofSMAD3(but not SMAD2)in the presence of TGF-b (Fig. 4D and fig.S8C). Overexpression of both BUB1 andTGFBRI induced a markedly greater TGF-b–stimulated interaction between BUB1and TGFBRI and further increased thephosphorylation of SMAD3 (Fig. 4D andfig. S8, B andC). These results demonstratethat overexpression of TGFBRI and/or BUB1enhances the activation of the signaling path-way only in the presence of the TGF-b ligand.Myc-tagged BUB1 coimmunoprecipitatedwith a His-tagged cytoplasmic tail of TGFBRIfrom A549 cell lysates (Fig. 4E), demon-strating that BUB1 may interact with thecytoplasmic tail of TGFBRI.

BUB1 mediates TGF-b–dependentTGFBRI-TGFBRII heteromericcomplex formationIn the presence of ligand, TGFBRI is re-cruited to the type II receptor (TGFBRII)

A

FL

BUB1

His

IB: FL

IB: His

siBUB1siTGFBRI

NSS

FL-SMAD3His-

-TGFBRI

IP:T

GFBR

IIn

put

B

FL

BUB1

His

IB: FL

IB: His

siBUB1siTGFBRI

NSS

FL-SMAD3His-

-

-TGFBRI

IP:T

GFBR

IIn

put

E

GAPDH

BUB1

Luciferase

siBUB1-

NSS

C

pSMAD2

pSMAD3

SMAD3

BUB1

GAPDH

SMAD2

TGFBRI

IB: SMAD4

SMAD4

IP:

SMAD

2/3

Inpu

t

siBUB1siTGFBRI

NSS

F

0

4

8

12

16

20

NSS siBUB1

SBE4

-Luc

(fol

d in

duct

ion)

**

0.0

0.4

0.8

1.2

1.6

2.0

NSS siBUB1

SBE4

-Luc

(fol

d in

duct

ion)

H

**

0

1

2

3

4

NSS siBUB1

SBE4

-Luc

(fol

d in

duct

ion)

G

**

D

0

2

4

6

8

10

NSS siBUB1

24 hours48 hours72 hours

SBE4

-Luc

(fol

d in

duct

ion)

****

**

- - - - - -

Fig. 2. BUB1 promotes the recruitment of SMAD3 to TGFBRI, SMAD2/3-SMAD4 complex formation, andtranscriptional response. (A and B) Immunoprecipitation (IP) for the TGFBRI antibody and then immuno-blotting (IB) for the Flag or His tags in HEK293T (A) and A549 (B) cells transfected with control siRNA(NSS), TGFBRI siRNA, or BUB1 siRNA along with Flag-tagged SMAD3 (FL-SMAD3) and 6XHis-taggedTGFBRI (His-TGFBRI) and treated with TGF-b (1 hour). (C) IP for SMAD2/3 followed by blotting for SMAD4in lysates from A549 cells transfected with control siRNA, TGFBR1 siRNA, or BUB1 siRNA and treated withTGF-b (1 hour). (D) Relative firefly luciferase activity (normalized toGaussia luciferase) after addition of TGF-b(10 ng/ml) and transfection withmock (NSS) or BUB1 siRNA for 24 to 72 hours in A549 cells transiently trans-fectedwith an SBE4-Luc reporter andGLuc plasmids. (E) Immunoblotting for luciferase andBUB1 in lysatesfrom A549 cells transfected and treated as in (D) and harvested 24 hours after TGF-b treatment. (F to H)Relative luciferase activity in NCI-H358 (F), MDA-231-1833 (G), and HeLa (H) transfected as in (D)and treated with TGF-b (10 ng/ml) for 24 hours. Blots are representative of three independent experiments.Data are means ± SEM of three independent experiments. **P < 0.001, two-sided Student’s t test.

R E S E A R C H A R T I C L E

www.SCIENCESIGNALING.org 6 January 2015 Vol 8 Issue 358 ra1 4

on January 6, 2015http://stke.sciencem

ag.org/D

ownloaded from

and forms a stable receptor complex throughtrans-phosphorylation (31). To examine therole for BUB1 in the formation of this het-eromeric receptor complex, A549 cells weretransfected with control or BUB1-targetedsiRNA and treated with TGF-b, and TGFBRIimmunoprecipitates from the resulting lysateswere probed for TGFBRII. BUB1 depletionabrogated the formation of the TGFBRI-TGFBRII complex (Fig. 4F). The abundanceof SARA [SMAD anchor for receptor activa-tion (32)] was detected in the immunoprecipi-tated fraction irrespective of BUB1 expression(Fig. 4F). TGF-b signaling is known to de-pend on the cell cycle (33), and BUB1 is aknown regulator of the cell cycle (8); however,we did not observe cell cycle arrest in BUB1-depleted cells (fig. S8D).

To test whether BUB1 kinase activityplayed a role in the formation of the hetero-meric receptor complex, we transfectedA549 cells with His-tagged TGFBRI fol-lowed by treatment with 2OH-BNPP1 for1 hour and stimulation with TGF-b for anadditional hour. The TGFBRI inhibitorSB431542 was used in parallel experimentsas a control. Inhibition of BUB1 kinase ac-tivity significantly reduced the formation ofthe TGFBRI-TGFBRII complex (Fig. 4Gand fig. S9). Inhibition of BUB1 kinase ac-tivity also decreased the coimmunoprecipi-tation ofMyc-taggedBUB1withHis-taggedTGFBRI (Fig. 4H). Because the kinase ac-tivity of BUB1 promoted the TGFBRI-TGFBRII complex formation and its owninteraction with TGFBRI, we investigatedwhether BUB1 could directly phosphorylateTGFBRI. In vitro kinase assays revealed thatBUB1 was not a direct kinase for TGFBRI(fig. S10), and although Myc-BUB1 andFlag-tagged SMAD2 coimmunoprecipi-tated when expressed in HEK293T cells(fig. S11), Myc-BUB1 did not phosphoryl-ate purified SMAD3 in vitro (fig. S12).Thus, albeit important, the target of BUB1’skinase activity in the TGFBRI signalingcomplex was elusive.

BUB1 interacts with TGFBRII inresponse to TGF-b stimulationWe next evaluated whether BUB1 also in-teracted with TGFBRII, which would beindicative of the formation of a ternary com-plex between BUB1, TGFBRI, and TGFBRII.HEK293T cells were transfected with His-TGFBRIandwild-typeMyc-BUB1and treatedwith 2OH-BNPP1 for 1 hour, followed bystimulation with TGF-b for an additionalhour. Coimmunoprecipitation assays revealedan increase in TGF-b–stimulated interaction

BUB1

GAPDH

Myc

pSMAD3

SMAD3

pSMAD2

SMAD2

****

C D E

H I

0

2

4

6

8

SBE4

-Luc

(fol

d in

duct

ion)

Mock Mock

pSMAD3

pSMAD2

SMAD2

Moc

k

pc-JUN

pP38MAPK

SMAD3

P38MAPK

pAKT

AKT

GAPDH

Moc

k

Moc

k

F

Moc

k

SB43

1542

2OH-

BNPP

1

IP:

SMAD

2/3

Inpu

t pSMAD3

pSMAD2

SMAD4

IB: SMAD4

SMAD2SMAD3

BUB1

GAPDH

G

Moc

k

SB43

1542

2OH-

BNPP

1

IP:

SMAD

2/3

Inpu

t SMAD2

pSMAD2

SMAD4

IB: SMAD4

pSMAD3

SMAD3

BUB1

GAPDH

0

5

10

15

20

25

SBE4

-Luc

(fol

d in

duct

ion)

A BsiBub1NSS

Myc-BUB1 WTVector

Myc-BUB1 KD

6

4

2

0

**

SBE4

-Luc

(fol

d in

duct

ion)

0.02

siBUB1NSS

Myc-BUB1 WTVector

Myc-BUB1 KD

TGFBRI

Moc

k

Moc

k

- (1 hour)

--

Fig. 3. BUB1 inhibitor 2OH-BNPP1 abrogates TGF-b signaling in a dose-dependent manner. (A) Im-munoblotting for total and phosphorylated (p) proteins as indicated in lysates from A549 cells tran-siently transfected with control siRNA (NSS) or BUB1 siRNA along with wild-type BUB1 (Myc-BUB1WT) or a kinase-deficient mutant (Myc-BUB1 KD), serum-starved, and then treated with TGF-b (10 ng/ml,1 hour). (B) Relative luciferase activity after addition of TGF-b in A549 cell cultures described in (A) expressingthe SBE4-Luc and GLuc plasmids. Data are means ± SEM of three independent experiments. (C to E)Immunoblots of lysates from A549 (C), NCI-H358 (D), and MDA-231-1833 (E) cells treated with vehicle,vehicle and TGF-b (10 ng/ml), or TGF-b and the indicated concentration of 2OH-BNPP1 for 1 hour. (F andG) IP for SMAD2/3 and then blotting for SMAD4 in lysates from A549 (F) and NCI-H358 (G) cells treatedwith vehicle, vehicle and TGF-b, or TGF-b and either SB431542 or 2OH-BNPP1 (10 mM) for 1 hour. (H and I)Relative luciferase activity in A549 (H) and NCI-H358 (I) cells treated as in (F) and (G) for 24 hours. Data aremeans ± SEM of three independent experiments. **P < 0.001, two-sided Student’s t test. Blots are represent-ative of three independent experiments; blots from (C) to (E) are quantified in fig. S7.

R E S E A R C H A R T I C L E

www.SCIENCESIGNALING.org 6 January 2015 Vol 8 Issue 358 ra1 5

on January 6, 2015http://stke.sciencem

ag.org/D

ownloaded from

between TGFBRII and BUB1 in HEK293T cells in a kinase-dependentmanner (Fig. 4I). In parallel studies, treatment of cells with the TGFBRI-specific inhibitor SB431542 did not affect this interaction (Fig. 4I).

BUB1 inhibitor 2OH-BNPP1 abrogates SMAD2phosphorylation in vivoWe followed up on our cellular studies by evaluating the abundance ofphosphorylated SMAD2 in subcutaneous A549 tumor xenograftsharvested 4 hours after mice were treated with 2OH-BNPP1. Similarto tumors from mice treated with the TGFBRI inhibitor, tumors from2OH-BNPP1–treated mice had a significantly decreased abundanceof phosphorylated SMAD2 compared with tumors from vehicle-

treated mice (Fig. 5, A and B), supporting a role for BUB1 in SMADactivation.

Together, our data indicate the formation of a ternary complex betweenTGFBRI, TGFBRII, and BUB1, and that the kinase activity of BUB1 maypromote canonical and noncanonical TGF-b signaling (Fig. 6).

DISCUSSION

Reverse genetic screens in mammalian cells using siRNA have been usedfor assessing gene function, target validation, pathway analysis, generedundancy, and therapeutic targeting of several disorders. In an effort toelucidate novel regulators of the TGF-b pathway, siRNA library screens

SARA

GAPDH

BUB1

TGFBRII

IB: TGFBRII

IB: SARA

siBUB1NSS

IP:

TGFB

RIIn

put

TGFBRI

D

SMAD3

His

pSMAD3

GAPDH

pSMAD2

SMAD2

IB: Myc

Myc

IP:

TGFB

RIIn

put

His-TGFBRIMyc-BUB1

Vector

E F

GAPDH

Myc

His

IB: Myc

IB: His

VectorHis-TGFBRI cytopls

IP: H

isIn

put

Myc-Bub1

BUB1 TGFBRI MergeM

ock

A

B C

G H

IB: TGFBRII

TGFBRII

SMAD2

IP: T

GFBR

IIn

put

IB: His

TGFBRI

pSMAD2

2OH-BNPP1SB431542

His-TGFBRI WTVector

SMAD3

pSMAD3

+–

++

+

––– ––– ––

+–

++ ++–– ––

–+

IP:T

GFBR

I

IB: Myc

IB: His

TGFBRI IPIgG IP

Inpu

t

His

SMAD3

SMAD2

Myc

pSMAD2

pSMAD3

His-TGFBRI WT + ++ ++Myc- BUB1 WT + ++ ++

2OH-BNPP1SB431542

+–+

+–– ––– ––

+––

+

TGFBRI

IB: TGFBRII

TGFBRII

SMAD2

IP: M

ycIn

put

IB: Myc

His

pSMAD2

SMAD3

pSMAD3

Myc

His-TGFBRI WT + ++ +Myc- Bub1 WT + ++ +

2OH-BNPP1SB431542

++

+–––

––

+–

–

+

–––

––

GAPDH

I

Mock

1 2 3 4 5 6 7 8

-

TGF-β (1 hour)

TGF-β (1 hour)

TGF- β (1 hour)

TGF- β (1 hour)TGF- β (1 hour)

Fig. 4. BUB1 colocalizes with TGFBRI, coim-munoprecipitates with TGFBRI and TGFBRII,and promotes heteromeric TGFBRI/II com-plex formation. (A) TIRF analysis of BUB1and TGFBRI colocalization in A549 cellstreated with TGF-b (10 ng/ml) for 72 hours.Scale bar, 10 mm. (B) Line scan across BUB1and TGFBRI particles within inset of (A).(C) Extent of colocalization normalized tothe mock sample. Data are means ± SEMof three independent experiments (1000particles, >20 cells). (D) IP for TGFBR1 andthen blotting for Myc in lysates from A549cells transfected with Myc-BUB1 and/orHis-TGFBRI, and treated with TGF-b. 1 hour:10 ng/ml for 1 hour. (E) IP for His and thenblotting for Myc in lysates from A549 cellstransfected with Myc-BUB1 and His-taggedcytosolic domain of TGFBRI. (F) IP for TGFBRIand then blotting for TGFBRII and SARA inlysates from A549 cells transfected withcontrol or BUB1 siRNA, serum-starved, andtreated with TGF-b. (G) IP for TGFBRI andthen blotting for His and TGFBRII in lysatesfrom A549 cells transfected with His-TGFBRI,serum-starved, and treated with SB431542or 2OH-BNPP1 (1 hour) and then TGF-b. (H)IP for TGFBRI and then blotting for Myc andTGFBRII in lysates from A549 cells transfectedwith His-TGFBRI and Myc-BUB1, serum-starved,and treated with SB431542 or 2OH-BNPP1(1 hour) and then TGF-b. (I) IP for Myc andthen blotting for TGFBRII in lysates fromHEK293T cells transfected with Myc-BUB1and His-TGFBRI, serum-starved, and treatedwith 2OH-BNPP1 or SB431542 (10 mM,1 hour) and then TGF-b. Blots are represent-ative of two (I) or at least three (D to H) in-dependent experiments.

R E S E A R C H A R T I C L E

www.SCIENCESIGNALING.org 6 January 2015 Vol 8 Issue 358 ra1 6

on January 6, 2015http://stke.sciencem

ag.org/D

ownloaded from

have identified key regulators of TGF-b signaling (34–40), although screensusing TGFBRI kinase activity as a readout have not been conducted. TheBTR reporter is unique in that loss of TGFBRI kinase activity results in again of function (increase in bioluminescence signal), and thus is less likelyto result in nonspecific or off-target biological effects than are commonlyobserved in traditional luciferase-based screens (4). Recent studies using aGFP (green fluorescent protein)–fused SMAD2 reporter in a large-scaleRNAi (RNA interference) screen observed that about 1% of hits exhibitedoff-target effects (41). In contrast, depletion of target genes that leads tocellular toxicity or off-target effects would be scored as false negative butnot false positive when using the BTR reporter. This provided us with con-fidence that siRNAs that resulted in an increase in reporter activity repre-sented true hits. A lack of overlap between hits derived from the two celllines is not surprising considering that the physiologic response of cells toTGF-b stimulation is often cell type–and context-dependent (42, 43). Wefocused on six hits that were common between the two cell lines, assumingthat these may represent important biological regulators. Of these, BMPR2,RPS6KB1, NEK6, and JAK2 have been described to be associated with theTGF-b pathway (44–48). However, the identification of BUB1 in our siRNAscreen using the TGFBRI kinase assay was surprising and unexpected dueto the preponderance of literature describing BUB1 as a component of themitotic machinery, wherein it ensures bipolar attachment to spindle micro-tubules so that chromatid segregation can occur with high fidelity (7, 8).Confirmatory studies conducted in normal cells (MCF10A, MRC5, and

Het1A) as well as tumor cell lines of various origins (A549, NCI-H358,MCF7, and MDA-231-1833) demonstrated a requirement for BUB1 ex-pression in the cellular response to TGF-b stimulation, and thus wasfollowed up for mechanistic studies. Additional confidence for BUB1 asa true hit was derived from studies demonstrating that depletion of BUB1in three independent cell lines resulted in abrogation of canonical (phos-phorylated SMAD2 and phosphorylated SMAD3) as well as noncanonical(phosphorylated p38 MAPK, AKT, and c-JUN) downstream effector mol-ecules. Because TGF-b signaling has been shown to be cell cycle–dependent(49, 50), we addressed the possibility that BUB1 (a regulator of mitosis)knockdown results in a decrease in TGF-b–mediated signaling due to the arrestof cells in mitosis. Existing evidence indicates that siRNA mediating com-plete depletion of BUB1 does not result in a cell cycle arrest (51), a findingconfirmed in our studies (fig. S9). In addition, cell cycle dependence ofTGF-b–mediated cellular responses is understood to be a downstream eventwherein changes in R-SMAD phosphorylation are not altered in a cellcycle–dependent manner but are rather regulated through interaction withSKI and SNON (5). Last, the ability of a BUB1 kinase inhibitor to abrogateTGF-b–dependent signaling within an hour of treatment, a period duringwhich a cell cycle arrest would not have occurred further, supports a directrole for BUB1 in receptor-mediated signaling.

The functional insights obtained in the current study suggest that BUB1is recruited to TGFBRI in a ligand-dependent manner. Coimmunoprecipi-tation studies demonstrated recruitment of BUB1 to TGFBRI within anhour of TGF-b treatment, whereas TIRF microscopy demonstrated coloca-lization only at the 72-hour, but not earlier, time point. Superior sensitivity

0

10

20

30

40

50

60

Mock SB431542 2OH-BNPP1

Treatments

n = 4 Mock n = 2 SB431542n = 5 2OH-BNPP1

****

A BMoMock (DMSck (DMSO)

SB431542 (10 mg/SB431542 (10 mg/kg)g)

Mock (DMSO)

2OH-BNPP1 (50 mg/2OH-BNPP1 (50 mg/kg)g)

SB431542 (10 mg/kg)

2OH-BNPP1 (50 mg/kg)

Fig. 5. BUB1 inhibitor blocks TGF-bsignaling in vivo. (A) Representativeimmunohistochemistry staining forphosphorylated SMAD2 in A549 xe-nografts harvested from SCID mice4 hours after treatment with 2OH-BNPP1

(50 mg/kg), SB431542 (10 mg/kg), or vehicle [dimethyl sulfoxide (DMSO)].Scale bar, 200 mm. (B) Number of cells staining positive for nuclear phos-phorylated SMAD2 in control (n = 4), SB431542-treated (n = 2), or 2OH-BNPP1–treated (n = 5) mice bearing tumors. Data are means ± SEM,cell counts from three random fields for each tumor. **P < 0.001, two-sidedStudent’s t test.

MEMBRANE

BUB1

EMT/migration/invasion

R-SMAD

P

R-SMAD

-

--

Fig. 6. A model for a role of BUB1 in TGF-b signaling. On the basis of ourfindings, we propose that BUB1 forms a ternary complex with TGFBRI andTGFBRII.Our data suggest that the interaction of BUB1with TGF-b receptorsis enhanced upon TGF-b stimulation, requires the kinase activity of BUB1,and promotes stabilization of the heteromeric complex between TGFBRI-TGFBRII, R-SMAD recruitment, and subsequent canonical and noncanonicalTGF-b signaling cascades.

R E S E A R C H A R T I C L E

www.SCIENCESIGNALING.org 6 January 2015 Vol 8 Issue 358 ra1 7

on January 6, 2015http://stke.sciencem

ag.org/D

ownloaded from

and signal to noise of the coimmunoprecipitation experiments may ex-plain this discordant observation. Alternatively, BUB1-TGFBRI inter-actions potentially occur at membrane-proximal regions, which areoptically inaccessible by TIRF microscopy, at earlier time points, andthese complexes are enriched at the plasma membrane only 72 hours afterligand treatment—a hypothesis that is still consistent with coimmuno-precipitation experiments that sample whole-cell lysates. Additionally,a requirement for BUB1 in the stabilization of the type II and I receptorcomplex,R-SMADrecruitment to the receptor, aswell as co-SMAD/R-SMADcomplex formation, and therefore in the downstream transcriptional re-sponse, was also demonstrated. Furthermore, our findings using a kinasedead (KD) mutant of BUB1 as well as a small-molecule inhibitor demon-strated that the kinase activity of BUB1 promotes the recruitment of BUB1to TGFBRI and TGFBRII, as well as for TGFBRI-TGFBRII complex for-mation upon ligand stimulation, leading to the propagation of the cellularresponse to TGF-b. The biological significance of BUB1’s role in TGF-bsignaling was emphasized by the finding that treatment of tumor-bearinganimals with 2OH-BNPP1 resulted in a loss of TGF-b–mediated SMAD2phosphorylation to an extent similar to SB431542, an inhibitor of TGFBRIkinase activity. Our in vitro kinase assays failed to demonstrate TGFBRI orSMAD3 as direct substrates of BUB1, although we observed an interactionbetween BUB1 and SMAD2 in HEK293T cells. Future studies will focuson identification of substrates of BUB1 kinase activity within the TGF-bsignaling pathway. Although we show that BUB1 forms a ternary complexby interacting with TGFBRI, as well as TGFBRII, the requirement forTGFBRII in the recruitment of BUB1 to TGFBRI was not directly investi-gated. Our studies in H358 cells, which lack TGFBRII (52–54), demonstrateda requirement of BUB1 expression and kinase activity in the propagation ofthe TGF-b signaling pathway, suggesting that TGFBRIImay be dispensablefor the interaction of BUB1 with TGFBRI.

In summary, we provide compelling evidence that BUB1 and its kinaseactivity are an integral component of TGF-b signaling and promoteTGFBRI/II receptor complex formation, thus regulating downstreamsignaling cascades, including the SMAD, MAPK, and PI3K/AKT path-ways. We further show that BUB1 interacts with both TGFBRI andTGFBRII and mediate TGF-b–dependent EMT, cell migration, and inva-sion. We anticipate that our results will lay the foundation for future studiesthat will provide new avenues for therapeutic targeting of TGF-b signalingin disease.

MATERIALS AND METHODS

Plasmid DNAWild-type BTR reporter has been described earlier (4). SBE4-Luc (6)reporter plasmid was provided by B. Vogelstein (Addgene plasmid #16527).pCMV5-TGFBRI-His (#19161), pCS2-Flag-SMAD2 (#14042), andpCS2-Flag-SMAD3 (#14052) were provided by J. Massague (55), andpCMV5B-TGFBRI-K232R (#11763) was a gift by J. Wrana (32). siRNA-resistant wild-type Myc-BUB1 and Myc-BUB1 KD have been previouslydescribed (27).

Cell culture and transfectionThe human lung carcinoma cell line A549 [American Type Culture collec-tion (ATCC)] was maintained in RPMI 1620 medium supplemented with10%heat-inactivated fetal bovine serum, 1%glutamine, and 0.1%penicillin/streptomycin/gentamicin (Gibco-Invitrogen). Cell cultures were grown in ahumidified incubator at 37°C and 5% CO2. Breast cancer cell line 1833 (56)derived fromMDA-MB-231 was provided by J. Massague (Memorial Slo-an Kettering Institute, NY) and maintained in Dulbecco’s modified Ea-

gle’s medium (DMEM) in the same conditions mentioned above. NCI-H358 lung cancer cell linewas provided byD. Beer (University ofMichigan,Ann Arbor). Normal lung fibroblast cell line MRC5, esophageal cell lineHet1A, breast epithelial lineMCF7, cervical cancer cell line HeLa, and normalkidney cell line HEK293Twere maintained in DMEM, whereas normalbreast epithelial cell line MCF10A was maintained in special medium.All cell lines were obtained from ATCC. Cell cultures were grown in a hu-midified incubator at 37°C and 5% CO2. A549 and MDA-231-1833 celllines stably expressing wild-type BTR and mutant reporter were generatedand maintained as described (4).

Antibodies, reagents, and siRNA libraryAntibodies to phosphorylated (pSer465/467) or total SMAD2, TGFBRI,TGFBRII, pSer63 c-JUN, pSer473 or total AKT, pThr180/Tyr182 or totalp38 MAPK, and GAPDH (glyceraldehyde-3-phosphate dehydrogenase)were all from Cell Signaling Technology. Antibody against SMAD3 wasfrom Invitrogen; antibodies to TGFBR1 (H100) and TGFBRII (L21) werefrom Santa Cruz; antibodies to Myc-tag and firefly luciferase were fromMillipore; antibodies to N terminus of BUB1 were from Abcam (rabbit)or Santa Cruz (goat); antibody to Flag–horseradish peroxidase (HRP)was from Sigma; and antibodies to His-HRP were from Invitrogen (clonesH3 and C-term) orMillipore (H8 clone). HRP- and fluorophore-conjugatedsecondary antibodies were from Jackson ImmunoResearch. Recombi-nant human TGF-b1 was obtained from HumanZyme. TGFBRI inhibi-tor SB431542 was obtained from Cayman Chemical. D-Luciferin wasfrom Xenogen Corp. The siGENOME SMARTpool siRNA library targetedagainst all human kinases, NSS, and individual siRNA from the siGENOMESMARTpool, andcustomsiRNAsagainstBUB1wereobtainedfromDharmacon.A specific, small-molecule inhibitor of BUB1 kinase activity (2-[(4-amino-1-(tert-butyl)-1H-pyrazolo[3,4-d]pyrimidin-3-yl)methyl]phenol; 2OH-BNPP1)(9) was synthesized in-house.

High-throughput siRNA screening againsthuman kinasesFor the high-throughput screen, A549 (5000 cells per well) andMDA-231-1833 cells (6000 cells per well) stably expressing the wild-type BTR repor-ter were plated in clear-bottom white-walled 96-well plates (Corning Inc.),1 day before transfection. Twenty microliters of 2 mM of each siRNA fromthe siRNA kinase library was added to a V-bottom intermediate plate. Simi-lar amount of control siRNA was also added in columns 1 and 12 in theintermediate plate (fig. S1A). The siRNAs were diluted by adding 40 mlof Opti-MEM and incubated for 5 min at room temperature. Eighty micro-liters of DharmaFECT 1 and Opti-MEMmixwas added to the intermediateplate and incubated for an additional 30 min. All liquid handling wasdone using Biomek NXP Laboratory Automation Workstation (BeckmanCoulter Inc.). The final assay concentration of each siRNA was 50 nM,whereas 0.2 ml of DF-1 was used for each well. Cells were incubated for72 hours in StoreX STX44 IC precision incubator (LiCONiC Instruments),which was connected to Plate Handler II robot (PerkinElmer) and EnVision2104 Plate Reader (PerkinElmer) with luciferin injector. TGF-b (10 ng/ml)was added to cells 1 hour before adding D-luciferin (50 mg/ml) andmeasuringbioluminescence. All the processes were automated. Fold change in reporteractivity was calculated over change in activity in nonsilencing control siRNA(NSS)–transfected cells.

To validate hits, 100 nM siRNA to selected genes, along with TGFBRIsiRNA andNSS, was transfected in A549, NCI-H358, andMDA-231-1833 cellsusing DharmaFECT 1. The transfected cells were incubated for 60 hoursand serum-starved overnight, and TGF-b (10 ng/ml) was added 1 hour be-fore harvesting cells for Western blotting for phosphorylated SMAD2/3,p38 MAPK, c-JUN, and AKT. Additionally, MRC5, Het1A, MCF7, and

R E S E A R C H A R T I C L E

www.SCIENCESIGNALING.org 6 January 2015 Vol 8 Issue 358 ra1 8

on January 6, 2015http://stke.sciencem

ag.org/D

ownloaded from

MCF10A cell lines were transfected in parallel experiments, and Westernblotting was done to probe for phosphorylated SMAD2.

Analysis of the siRNA screenData from the siRNA screen were initially analyzed on MS Excel. Aquartile-based method was used for HTS (high-throughput screening) hitselection (20). Normalization was done with a control nontargeting siRNA(NSS) placed on each individual assay plate, and median (Q2), first (Q1),and third (Q3) quartile values were calculated for all normalized valuesand subjected to plate-by-plate analysis. To facilitate experiment-wideanalysis, Q1, Q2, and Q3 values were calculated as described above.Averages and SE calculations were done for triplicates of log(x/median)values. Upper and lower boundaries were calculated as Q3 + 2c(Q3 − Q2)andQ1−2c(Q2−Q1), respectively, forc=1.7239, corresponding to a high-stringency targeted error rate (a = 0.0027; more than 7.14-fold activation inA549 cells and more than 13.91-fold in MDA-231-1833 cells), and c =0.9826, corresponding to a lower-stringency targeted error rate (a = 0.046;more than5.4-fold inA549andmore than10.29-fold inMDA-231-1833cells).High-stringency hits were chosen as those targets that scored a ≤ 0.0027 inboth plate-by-plate and experiment-wide analyses. Selected targets distrib-uted equally among all the plates, which indicates that our data did not showany “alphabetical clustering,” rather agreeing on the sparse-hit hypothesis,which assumes that hits are randomly spaced throughout the data set.Selected top hits were subjected to Pathway Analysis using PathwayGuide(Advaita Corp.) to identify the pathways that are significantly affected inthe BTR assay. This software uses an impact analysis that includes the clas-sical statistics and also considers the magnitude of each gene’s expressionchange, their type and position in the given pathways, their interaction, etc.(57, 58). Two-way evidence plot, total perturbation accumulation (TA)plot, and pathway maps were generated by PathwayGuide. Functionalprotein association network analysis of the hits was carried out using String(http://string-db.org/). Heatmaps of low- and high-stringency hitswere gen-erated using tMeV (59) to show data reproducibility.

Western blot analysisWestern analysis was carried out using standard protocols. Cells were grownin culture dishes, transfected with specific siRNA, or treated with selectcompounds for designated time periods, and cell lysates were resolved onSDS–polyacrylamide gel electrophoresis (SDS-PAGE) gels and transferredto polyvinylidene difluoride (PVDF) membranes. Membranes were probedagainst specific primary antibodies followed by HRP-conjugated secondaryantibodies and thenvisualized using the enhanced chemiluminescence (ECL)Western Blotting System (GE Healthcare). Signal intensity was measuredusing an image processing and analysis program (ImageJ, v1.45) (60).

CoimmunoprecipitationFor coimmunoprecipitation studies, HEK293T or A549 cells were trans-fected with various plasmids. Lysates were made 48 hours after transfectionin native lysis buffer [50 mM tris (pH 7.4), 1%NP-40, 0.25% deoxycholatesodium salt, 150mMNaCl, 10%glycerol, and 1mMEDTA] supplementedwith 1× PhosSTOP (Roche), 1× Protease Inhibitor Cocktail (Roche), sodiumorthovanadate, sodium fluoride, phenylmethylsulfonyl fluoride (PMSF),and b-glycerol phosphate (2 mM each). Cell pellet was solubilized in 600 mlof lysis buffer, rocked for 60 min at 4°C, and centrifuged at 14,000 rpm for15 min. Protein estimation was performed with detergent-compatible DCAssay Kit (Pierce). Lysates were precleared by incubating with 50 ml ofagarose beads for 1 hour at 4°C and centrifuged. Coimmunoprecipitationwas carried out by incubating precleared cell lysate (400 mg of protein) with1 to 1.5 mg of specific antibody overnight at 4°C. The immune complexwascaptured using 30 ml of slurry of proteinA/G–coupled Sepharose beads (GE

Healthcare) for 2 to 3 hours and then washed four times with lysis buffer.The resulting pellet was resolved by SDS-PAGE and transferred to PVDFmembrane for Western analysis.

SBE4-Luc reporter assayForSBE4-Luc reporter studies, 1.25×105 cells (A549,NCI-H358,MDA-231-1833, and HeLa) were plated 1 day before transfection. Cells were transfectedthe next day with 100 ng of SBE4-Luc plasmid, 100 nM siRNA, 500 ng ofvariousDNA, and20ngofGaussia luciferase for eachwellwithLipofectamine2000. Cells were serum-starved 24 hours after transfection for 6 hours andtreatedwithTGF-b (10ng/ml) for 24hours (or up to72hours forA549) beforemeasuring SBE4-Luc bioluminescence activity. Alternately, cells were treatedwith 10 mMSB431542 or 2OH-BNPP1 in the presence of TGF-b. Cells werewashed twice with phosphate-buffered saline (PBS) and kept in fresh me-dium for 8 to 24 hours beforemeasuringGaussia luciferase activity.Relativefirefly luciferase activity was normalized to Gaussia luciferase activity.

Cell migration and invasion assayMultiwell cell culture inserts (8.0-mm pore size, BD Bioscience) were usedfor in vitro migration assays, whereas Matrigel precoated inserts (8.0-mmpore size, BD Bioscience) were used for invasion assays. A549 cells weretransfected with 100 nM siRNA, serum-starved for 6 hours with 0.5% fetalbovine serum–containing medium, treated with TGF-b (10 ng/ml) or mockfor 72 hours, trypsinized, washed, and plated in serum-free medium in theinserts (25K for migration, 50K for invasion). Medium with 5% serum wasadded in the bottom chamber, providing a chemotactic gradient. Plateswereincubated for 8 hours (migration assay) or for 24 hours (invasion assay). Atthe end of the assay, cells were removed from the top side of the insert usinga cotton swab. Cells that penetrated to the underside surface were fixedusing absolute methanol, stained with 0.5% crystal violet (Sigma), andcounted under the microscope. The assay was repeated three times. Themean of three repeats run in triplicates is plotted.

ImmunofluorescenceA549 cells were plated on glass coverslips in six-well plates and transfectedwith 100nMsiRNA, serum-starvedovernight, treatedwithTGF-b (10ng/ml)for 1 hour, and processed for immunofluorescence. For the inhibitor studies,cells were plated on glass coverslips, serum-starved for 24 hours, and treatedwith TGF-b (10 ng/ml) in the presence of 10 mM inhibitors (SB431542 or2OH-BNPP1) for 1 hour. Cells on coverslips were washed with PBS andfixed in 10% (v/v) buffered formalin (5 min at room temperature), followedby methanol fixation at −20°C for 20 min. Coverslips were rehydrated byincubating three times in PBS for 5 min and permeabilized in 0.2% (v/v) Tri-tonX-100–PBS containing 1% (w/v) bovine serumalbumin (BSA) on ice for5min. Coverslips were blocked by 5% (w/v) BSA, 5% (v/v) goat serum, and10% (v/v) donkey serum in PBS for 1 hour. Anti-SMAD2 antibody at 1:200dilution was added in PBS containing 0.5% (w/v) BSA and incubated for1 hour at 37°C in a humidified chamber. Cover glasses were washed threetimes in PBS containing 0.025% (v/v) Triton X-100 and incubated with anti-rabbit Alexa Fluor 488 secondary antibody for 1 hour. After washing,coverslips were mounted in ProLong Gold (Invitrogen) containing DAPI(4′,6-diamidino-2-phenylindole), dried overnight at room temperature, andstored at −20°C until microscopy. Micrographs were taken using OlympusBX51 upright fluorescence microscope fitted with an Olympus DP70 high-resolution digital color camera.

Tumor xenograft experimentsand immunohistochemistryCell line–derived xenograftswere established by implanting 2.5× 106A549cells subcutaneously into each flank of 4- to 6-week-old male SCID mice.

R E S E A R C H A R T I C L E

www.SCIENCESIGNALING.org 6 January 2015 Vol 8 Issue 358 ra1 9

on January 6, 2015http://stke.sciencem

ag.org/D

ownloaded from

When tumors reached a volume between 40 and 60mm3, micewere injectedwith a single intraperitoneal dose of SB431542 (10 mg/kg body weight), adose previously reported to inhibit TGF-b signaling in mouse models (61),2OH-BNPP1 (50 mg/kg), or vehicle (DMSO). Tumors were excised 4 hoursafter treatment and fixed in formalin. Paraffin-embedded sections werestainedusing anantibody for phosphorylatedSMAD2, andmicrographsweretaken with an Olympus microscope fitted with an Olympus DP70 high-resolution digital camera. The proportion of nuclei that stained positive forphosphorylated SMAD2 were counted in three random fields per tumorper treatment condition [vehicle (n= 4), SB431542 (n= 2), and 2OH-BNPP1(n = 5)]. A two-sided Student’s t test was performed to assess statistical sig-nificance. Slideswere adjusted for brightness and contrast withAdobe Photo-shop CS2 (Adobe Inc.), but the micrographs underwent no furthermanipulations. All mouse experiments were approved by the UniversityCommittee on the Use and Care of Animals of the University of Michigan.

Cell cycle analysisA549 cellswere plated in six-well plates and transfectedwith 100 nMsiRNA.Cells were trypsinized 72 hours after transfection, counted, resuspended inPBS (50,000 cells/ml), fixed using 50% ethanol, stained with propidium io-dide (50mg/ml) supplementedwith ribonucleaseA (100mg/ml), and readon aflow cytometer. The mean of at least three experiments is plotted.

TIRF microscopyA549 cellswere plated onglass coverslip bottom35-mmdishes (MatTek) andserum-starved the next day for 24 hours. Cells were treated with TGF-b (10ng/ml) for 1, 24, or 72 hours and washed with PBS. Cells were fixed in 10%(v/v) buffered formalin for 5min at room temperature, followed bymethanolfixation at −20°C for 20 min. Cells were rehydrated by incubating threetimes in PBS for 5 min and were permeabilized in 0.2% (v/v) TritonX-100–PBS containing 1%BSA on ice for 5min. Cells were blocked using5% (w/v) BSA, 5% (v/v) goat serum, and 10% (v/v) donkey serum in PBSfor 1 hour. Primary antibodies at appropriate dilution [anti-TGFBRI, 1:500;anti-BUB1 (goat), 1:35] were added in PBS containing 0.5% (w/v) BSAand incubated for 1 hour at room temperature in a humidified chamber. Plateswere washed three times in PBS containing 0.025% (v/v) Triton X-100 andincubatedwith fluorophore-labeled secondary antibodies for 1 hour in the dark.After washing, fresh PBSwas added into the plates, and TIRFmicroscopywasperformed at the Single Molecule Analysis in Real-Time (SMART) Centerusing a cell-TIRF system based on an Olympus IX81 microscope, equippedwith a 60× 1.49 NA (numerical aperture) oil immersion objective (Olympus).Samples were placed on a nanometer-precision motorized stage (PrincetonInstruments) with xyz position piezo control. Imageswere acquired on an iXonUltra EMCCD (electronmultiplying charge-coupled device) camera (512 ×512 pixels; Andor). Alexa Fluor 488 and Alexa Fluor 647 dyes conjugatedto the secondary antibodies were excited, one after another, by the 488- and640-nm lasers, respectively. Their fluorescencewas filtered through a cus-tomized quad-band filter cube consisting of a ZT405/488/561/640rpc di-chroic filter (Chroma) and a ZET405/488/561/640m emission filter(Chroma). Images of BUB1 (Alexa Fluor 488) and TGFBRI (Alexa Fluor647) were processed using a custom-compiled ImageJ macro to identify spa-tially distinct particles and calculate extent of colocalization.

In vitro kinase assayInvitro kinase assayswereperformedwith200ngof protein (wild-typeBUB1ora kinase inactive mutant BUB1 KD) and 2 mg of substrate (wild-type TGFBRI,TGFBRIKD,SMAD3,orH2A) in1×kinasebuffer [50mMtris-HCl, 150mMNaCl, 10 mM MgCl2, 1% (v/v) glycerol, 0.1% (v/v) Triton X-100, DTT(dithiothreitol), PMSF, Na3VO4 (1 mM each), 2 mM NaF, and b-glycerolphosphate] in 20-ml volume containing 10 mCi [32P]ATP (adenosine

5′-triphosphate) and 300 mM cold ATP. Reactions were run at 30°C for0.5 to 1 hour, quenched using Laemmli buffer, and resolved using a 4 to12% bis-tris gel. Quantitative autoradiography was performed with a Ty-phoon FLA 9000 scanner (GE Healthcare).

SUPPLEMENTARY MATERIALSwww.sciencesignaling.org/cgi/content/full/8/358/ra1/DC1Fig. S1. siRNA screening and pathway analysis.Fig. S2. BUB1 mediates TGF-b–dependent SMAD2/3 phosphorylation.Fig. S3. Depletion of BUB1 leads to reduced SMAD3 recruitment to TGFBRI and TGF-b–dependent accumulation of SMAD2 in the nucleus.Fig. S4. Depletion of BUB1 attenuates TGF-b–dependent SMAD2/3-SMAD4 complex formationin A549 cells.Fig. S5. Depletion of BUB1 inhibits TGF-b–mediated EMT in A549 and NCI-H358 cells.Fig. S6. Depletion of BUB1 inhibits TGF-b–mediated migration and invasion of A549 cells.Fig. S7. BUB1 kinase activity mediates TGF-b–dependent phosphorylation and nuclearactivity of R-SMAD.Fig. S8. TGFBRI and BUB1 colocalization by TIRF microscopy at 1 and 24 hours andcoimmunoprecipitation of His-TGFBRI and Myc-BUB1.Fig. S9. BUB1 kinase activity mediates TGFBRI-TGFBRII interaction.Fig. S10. TGFBRI is not a direct substrate of BUB1 kinase activity.Fig. S11. Wild-type Myc-BUB1 interacts with FL-SMAD2 in HEK293T cells.Fig. S12. SMAD3 is not a direct substrate of BUB1 kinase activity.Table S1. Human kinome siRNA screen in A549-BTR and MDA-231-1833–BTR cells.Table S2. Hits obtained in A549-BTR and MDA-231-1833–BTR human kinome screen.Table S3. Pathway impact analysis based on the fold induction of the BTR reporter.

REFERENCES AND NOTES1. Y. Drabsch, P. ten Dijke, TGF-b signalling and its role in cancer progression and me-

tastasis. Cancer Metastasis Rev. 31, 553–568 (2012).2. J. Xu, S. Lamouille, R. Derynck, TGF-b-induced epithelial to mesenchymal transition.

Cell Res. 19, 156–172 (2009).3. R. Derynck, R. J. Akhurst, A. Balmain, TGF-b signaling in tumor suppression and

cancer progression. Nat. Genet. 29, 117–129 (2001).4. S. Nyati, K. Schinske, D. Ray, M. Nyati, B. D. Ross, A. Rehemtulla, Molecular imaging

of TGFb-induced Smad2/3 phosphorylation reveals a role for receptor tyrosine ki-nases in modulating TGFb signaling. Clin. Cancer Res. 17, 7424–7439 (2011).

5. A. Zieba, K. Pardali, O. Söderberg, L. Lindbom, E. Nyström, A. Moustakas, C. H. Heldin,U. Landegren, Intercellular variation in signaling through the TGF-b pathway and itsrelation to cell density and cell cycle phase.Mol. Cell. Proteomics 11, M111.013482 (2012).

6. L. Zawel, J. L. Dai, P. Buckhaults, S. B. Zhou, K. W. Kinzler, B. Vogelstein, S. E. Kern,Human Smad3 and Smad4 are sequence-specific transcription activators. Mol. Cell 1,611–617 (1998).

7. V. L. Johnson, M. I. F. Scott, S. V. Holt, D. Hussein, S. S. Taylor, Bub1 is required forkinetochore localization of BubR1, Cenp-E, Cenp-F and Mad2, and chromosome con-gression. J. Cell Sci. 117, 1577–1589 (2004).

8. H. Yu, Z. Tang, Bub1 multitasking in mitosis. Cell Cycle 4, 262–265 (2005).9. J. S. Kang, M. J. Yang, B. Li, W. Qi, C. Zhang, K. M. Shokat, D. R. Tomchick, M. Machius,

H. T. Yu, Structure and substrate recruitment of the human spindle checkpoint kinaseBub1. Mol. Cell 32, 394–405 (2008).

10. S. Lawler, X. H. Feng, R. H. Chen, E. M. Maruoka, C. W. Turck, I. Griswold-Prenner,R. Derynck, The type II transforming growth factor-b receptor autophosphorylates notonly on serine and threonine but also on tyrosine residues. J. Biol. Chem. 272,14850–14859 (1997).

11. S. Souchelnytskyi, P. ten Dijke, K. Miyazono, C. H. Heldin, Phosphorylation of Ser165in TGF-beta type I receptor modulates TGF-beta1-induced cellular responses. EMBOJ. 15, 6231–6240 (1996).

12. A. J. Galliher, W. P. Schiemann, Src phosphorylates Tyr284 in TGF-b type II receptorand regulates TGF-b stimulation of p38 MAPK during breast cancer cell proliferationand invasion. Cancer Res. 67, 3752–3758 (2007).

13. P. Kavsak, R. K. Rasmussen, C. G. Causing, S. Bonni, H. Zhu, G. H. Thomsen,J. L. Wrana, Smad7 binds to Smurf2 to form an E3 ubiquitin ligase that targets the TGFbreceptor for degradation. Mol. Cell 6, 1365–1375 (2000).

14. J. S. Kang, E. F. Saunier, R. J. Akhurst, R. Derynck, The type I TGF-b receptor iscovalently modified and regulated by sumoylation. Nat. Cell Biol. 10, 654–664 (2008).

15. W. Shi, C. Sun, B. He, W. Xiong, X. Shi, D. Yao, X. Cao, GADD34–PP1c recruited bySmad7 dephosphorylates TGFb type I receptor. J. Cell Biol. 164, 291–300 (2004).

16. F. Huang, Y. G. Chen, Regulation of TGF-b receptor activity. Cell Biosci. 2, 9 (2012).17. J. S. Kang, C. Liu, R. Derynck, New regulatory mechanisms of TGF-b receptor

function. Trends Cell Biol. 19, 385–394 (2009).18. C. Le Roy, J. L. Wrana, Clathrin- and non-clathrin-mediated endocytic regulation of

cell signalling. Nat. Rev. Mol. Cell Biol. 6, 112–126 (2005).

R E S E A R C H A R T I C L E

www.SCIENCESIGNALING.org 6 January 2015 Vol 8 Issue 358 ra1 10

on January 6, 2015http://stke.sciencem

ag.org/D

ownloaded from

19. H. Mitchell, A. Choudhury, R. E. Pagano, E. B. Leof, Ligand-dependent and -independenttransforming growth factor-b receptor recycling regulated by clathrin-mediated endo-cytosis and Rab11. Mol. Biol. Cell 15, 4166–4178 (2004).

20. X. H. D. Zhang, X. T. C. Yang, N. J. Chung, A. Gates, E. Stec, P. Kunapuli, D. J. Holder,M. Ferrer, A. S. Espeseth, Robust statistical methods for hit selection in RNA interferencehigh-throughput screening experiments. Pharmacogenomics 7, 299–309 (2006).

21. M. M. Martin, J. A. Buckenberger, J. Jiang, G. E. Malana, D. L. Knoell, D. S. Feldman,T. S. Elton, TGF-b1 stimulates human AT1 receptor expression in lung fibroblasts bycross talk between the Smad, p38 MAPK, JNK, and PI3K signaling pathways. Am. J.Physiol. Lung Cell. Mol. Physiol. 293, L790–L799 (2007).

22. J. Y. Yi, I. Shin, C. L. Arteaga, Type I transforming growth factor b receptor binds toand activates phosphatidylinositol 3-kinase. J. Biol. Chem. 280, 10870–10876 (2005).

23. J. P. Thiery, J. P. Sleeman, Complex networks orchestrate epithelial–mesenchymaltransitions. Nat. Rev. Mol. Cell Biol. 7, 131–142 (2006).

24. E. Meulmeester, P. Ten Dijke, The dynamic roles of TGF-b in cancer. J. Pathol. 223,205–218 (2011).

25. H. Kasai, J. T. Allen, R. M. Mason, T. Kamimura, Z. Zhang, TGF-b 1 induces humanalveolar epithelial to mesenchymal cell transition (EMT). Respir. Res. 6, 56 (2005).

26. J. H. Kim, Y. S. Jang, K. S. Eom, Y. I. Hwang, H. R. Kang, S. H. Jang, C. H. Kim, Y. B. Park,M. G. Lee, I. G. Hyun, K. S. Jung, D. G. Kim, Transforming growth factor b1 inducesepithelial-to-mesenchymal transition of A549 cells. J. Korean Med. Sci. 22, 898–904 (2007).

27. W. Qi, H. T. Yu, KEN-box-dependent degradation of the Bub1 spindle checkpoint kinaseby the anaphase-promoting complex/cyclosome. J. Biol. Chem. 282, 3672–3679 (2007).

28. G. J. Inman, F. J. Nicolás, J. F. Callahan, J. D. Harling, L. M. Gaster, A. D. Reith, N. J. Laping,C. S. Hill, SB-431542 is a potent and specific inhibitor of transforming growth factor-bsuperfamily type I activin receptor-like kinase (ALK) receptors ALK4, ALK5, and ALK7.Mol. Pharmacol. 62, 65–74 (2002).

29. D. Perera, V. Tilston, J. A. Hopwood, M. Barchi, R. P. Boot-Handford, S. S. Taylor,Bub1 maintains centromeric cohesion by activation of the spindle checkpoint. Dev.Cell 13, 566–579 (2007).

30. S. Pitchiaya, J. R. Androsavich, N. G. Walter, Intracellular single molecule microscopyreveals two kinetically distinct pathways for microRNA assembly. EMBO Rep. 13,709–715 (2012).

31. M. Kawabata, A. Chytil, H. L. Moses, Cloning of a novel type II serine/threonine ki-nase receptor through interaction with the type I transforming growth factor-b receptor.J. Biol. Chem. 270, 5625–5630 (1995).

32. T. Tsukazaki, T. A. Chiang, A. F. Davison, L. Attisano, J. L. Wrana, SARA, a FYVEdomain protein that recruits Smad2 to the TGFb receptor. Cell 95, 779–791 (1998).

33. Y. Wan, X. Liu, M. W. Kirschner, The anaphase-promoting complex mediates TGF-bsignaling by targeting SnoN for destruction. Mol. Cell 8, 1027–1039 (2001).

34. P. Cherukuri, A. J. DeCastro, A. L. Balboni, S. L. Downey, J. Y. Liu, J. A. Hutchinson,J. DiRenzo, Phosphorylation of DNp63a via a novel TGFb/ALK5 signaling mechanismmediates the anti-clonogenic effects of TGFb. PLOS One 7, e50066 (2012).

35. D. J. de Gorter, M. van Dinther, O. Korchynskyi, P. ten Dijke, Biphasic effects oftransforming growth factor b on bone morphogenetic protein–induced osteoblast dif-ferentiation. J. Bone Miner. Res. 26, 1178–1187 (2011).

36. S. Dupont, A. Mamidi, M. Cordenonsi, M. Montagner, L. Zacchigna, M. Adorno, G. Martello,M. J. Stinchfield, S. Soligo, L. Morsut, M. Inui, S. Moro, N. Modena, F. Argenton,S. J. Newfeld, S. Piccolo, FAM/USP9x, a deubiquitinating enzyme essential for TGFbsignaling, controls Smad4 monoubiquitination. Cell 136, 123–135 (2009).

37. J. Izrailit, H. K. Berman, A. Datti, J. L. Wrana, M. Reedijk, High throughput kinaseinhibitor screens reveal TRB3 and MAPK-ERK/TGFb pathways as fundamental Notchregulators in breast cancer. Proc. Natl. Acad. Sci. U.S.A. 110, 1714–1719 (2013).

38. J. Mullenders, A. W. Fabius, M. M. van Dongen, H. J. Kuiken, R. L. Beijersbergen,R. Bernards, Interleukin-1R–associated kinase 2 is a novel modulator of the transforminggrowth factor b signaling cascade. Mol. Cancer Res. 8, 592–603 (2010).

39. H. Yu, M. Königshoff, A. Jayachandran, D. Handley, W. Seeger, N. Kaminski, O. Eickelberg,Transgelin is a direct target of TGF-b/Smad3-dependent epithelial cell migration in lungfibrosis. FASEB J. 22, 1778–1789 (2008).

40. L. Zhang, F. F. Zhou, Y. Drabsch, R. Gao, B. E. Snaar-Jagalska, C. Mickanin, H. Z. Huang,K. A. Sheppard, J. A. Porter, C. X. Lu, P. ten Dijke, USP4 is regulated by AKT phospho-rylation and directly deubiquitylates TGF-b type I receptor.Nat. Cell Biol. 14, 717–726 (2012).

41. N. Schultz, D. R. Marenstein, D. A. De Angelis, W. Q. Wang, S. Nelander, A. Jacobsen,D. S. Marks, J. Massagué, C. Sander, Off-target effects dominate a large-scale RNAiscreen for modulators of the TGF-b pathway and reveal microRNA regulation ofTGFBR2. Silence 2, 3 (2011).

42. H. Ikushima, K. Miyazono, Cellular context-dependent “colors” of transforming growthfactor-b signaling. Cancer Sci. 101, 306–312 (2010).

43. S. S. Söderberg, G. Karlsson, S. Karlsson, Complex and context dependent regulation ofhematopoiesis by TGF-b superfamily signaling. Ann. N. Y. Acad. Sci. 1176, 55–69 (2009).

44. C. Petritsch, H. Beug, A. Balmain, M. Oft, TGF-b inhibits p70 S6 kinase via proteinphosphatase 2A to induce G1 arrest. Genes Dev. 14, 3093–3101 (2000).

45. P. ten Dijke, H. Yamashita, H. Ichijo, P. Franzén, M. Laiho, K. Miyazono, C. H. Heldin,Characterization of type I receptors for transforming growth factor-beta and activin.Science 264, 101–104 (1994).

46. M. Barrios-Rodiles, K. R. Brown, B. Ozdamar, R. Bose, Z. Liu, R. S. Donovan, F. Shinjo,Y. M. Liu, J. Dembowy, I. W. Taylor, V. Luga, N. Przulj, M. Robinson, H. Suzuki,Y. Hayashizaki, I. Jurisica, J. L. Wrana, High-throughput mapping of a dynamic signalingnetwork in mammalian cells. Science 307, 1621–1625 (2005).

47. C. Dees, M. Tomcik, K. Palumbo, A. Akhmetshina, A. Horn, P. Zerr, O. Distler, G. Schett,J. H. W. Distler, JAK2 mediates the stimulatory effects of transforming growth factor betaon fibroblast activation and tissue fibrosis. Arthritis Rheum. 63, S944–S945 (2011).

48. X. Wang, S. Shaw, F. Amiri, D. C. Eaton, M. B. Marrero, Inhibition of the Jak/STATsignaling pathway prevents the high glucose-induced increase in TGF-b and fibronec-tin synthesis in mesangial cells. Diabetes 51, 3505–3509 (2002).

49. J. Song, EMT or apoptosis: A decision for TGF-b. Cell Res. 17, 289–290 (2007).50. T. Hirschhorn, L. Barizilay, N. I. Smorodinsky, M. Ehrlich, Differential regulation of

Smad3 and of the type II transforming growth factor-b receptor in mitosis: Implicationsfor signaling. PLOS One 7, e43459 (2012).

51. P. Meraldi, P. K. Sorger, A dual role for Bub1 in the spindle checkpoint and chromo-some congression. EMBO J. 24, 1621–1633 (2005).

52. T. K. Kim, E. K. Mo, C. G. Yoo, C. T. Lee, S. K. Han, Y. S. Shim, Y. W. Kim, Alter-ation of cell growth and morphology by overexpression of transforming growth factorb type II receptor in human lung adenocarcinoma cells. Lung Cancer 31, 181–191(2001).

53. W. S. Kim, C. Park, Y. S. Jung, H. S. Kim, J. Han, C. H. Park, K. Kim, J. Kim, Y. M. Shim,K. Park, Reduced transforming growth factor-beta type II receptor (TGF-beta RII) expres-sion in adenocarcinoma of the lung. Anticancer Res. 19, 301–306 (1999).

54. Y. Shintani, M. Maeda, N. Chaika, K. R. Johnson, M. J. Wheelock, Collagen I pro-motes epithelial-to-mesenchymal transition in lung cancer cells via transforminggrowth factor-b signaling. Am. J. Respir. Cell Mol. Biol. 38, 95–104 (2008).

55. A. Hata, R. S. Lo, D. Wotton, G. Lagna, J. Massagué, Mutations increasing auto-inhibition inactivate tumour suppressors Smad2 and Smad4. Nature 388, 82–87(1997).

56. Y. B. Kang, P. M. Siegel, W. P. Shu, M. Drobnjak, S. M. Kakonen, C. Cordón-Cardo,T. A. Guise, J. Massagué, A multigenic program mediating breast cancer metastasisto bone. Cancer Cell 3, 537–549 (2003).

57. S. Draghici, P. Khatri, A. L. Tarca, K. Amin, A. Done, C. Voichita, C. Georgescu, R. Romero,A systems biology approach for pathway level analysis. Genome Res. 17, 1537–1545(2007).

58. A. L. Tarca,S.Draghici, P. Khatri, S.S.Hassan,P.Mittal, J.S.Kim,C. J. Kim, J. P. Kusanovic,R. Romero, A novel signaling pathway impact analysis. Bioinformatics 25, 75–82(2009).

59. A. I. Saeed, V. Sharov, J. White, J. Li, W. Liang, N. Bhagabati, J. Braisted, M. Klapa,T. Currier, M. Thiagarajan, A. Sturn, M. Snuffin, A. Rezantsev, D. Popov, A. Ryltsov,E. Kostukovich, I. Borisovsky, Z. Liu, A. Vinsavich, V. Trush, J. Quackenbush, TM4: A free,open-source system for microarray data management and analysis. Biotechniques34, 374–378 (2003).

60. M. D. Abramoff, P. J. Magalhães, S. J. Ram, Image processing with ImageJ. BiophotonicsInt. 11, 36–42 (2004).

61. Y. Chen, C. S. Kam, F. Q. Liu, Y. Liu, V. C. Lui, J. R. Lamb, P. K. Tam, LPS-induced up-regulation of TGF-b receptor 1 is associated with TNF-a expression in human monocyte-derived macrophages. J. Leukoc. Biol. 83, 1165–1173 (2008).

Acknowledgments: We thank S. Kronenberg for help in figure graphics, and T. Cunninghamwith preparation of the manuscript. We also thank the Microscopy & Image Analysis Laboratoryand the DNA Sequencing Core at the University of Michigan Medical School. Funding: Thiswork was supported by grants from the NIH: P01CA85878, P50CA093990, and R01CA136892(B.D.R. and A.R.), P50CA093990-11 (Career Development Award; S.N.), as well as NSF MRI-R2-ID award DBI-0959823 (N.G.W.). Author contributions: S.N., B.D.R., and A.R. conceivedand designed the study and wrote the manuscript with input from all the authors. S.N. performedmost of the experiments. K.S.-S. assisted in the high-throughput screen. S.P. performed TIRFmicroscopy analysis. K.C., A.C., N.C., and J.D. provided experiment support. M.E.V.D. synthe-sized the 2OH-BNPP1 inhibitor. S.V. and C.K.-S. assisted, and N.G.W., M.K.N., and D.R. ana-lyzed the HTS data. H.Y. provided reagents, contributed to the experimental design, andassisted in the interpretation of results. Competing interests: The authors declare that theyhave no competing interests. Data and materials availability: Data for the kinome siRNAscreen are deposited in PubChem (https://pubchem.ncbi.nlm.nih.gov/assay/assay.cgi?aid=1117269), accession ID 1117269. Materials may be requested from the corresponding au-thor. There is a patent pending for the use of targeting BUB1 therapeutically.

Submitted 14 April 2014Accepted 7 November 2014Final Publication 6 January 201510.1126/scisignal.2005379Citation: S. Nyati, K. Schinske-Sebolt, S. Pitchiaya, K. Chekhovskiy, A. Chator,N. Chaudhry, J. Dosch, M. E. Van Dort, S. Varambally, C. Kumar-Sinha, M. K. Nyati,D. Ray, N. G. Walter, H. Yu, B. D. Ross, A. Rehemtulla, The kinase activity of the Ser/Thrkinase BUB1 promotes TGF-b signaling. Sci. Signal. 8, ra1 (2015).

R E S E A R C H A R T I C L E

www.SCIENCESIGNALING.org 6 January 2015 Vol 8 Issue 358 ra1 11

on January 6, 2015http://stke.sciencem

ag.org/D

ownloaded from

(358), ra1. [doi: 10.1126/scisignal.2005379]8Science Signaling (January 6, 2015) G. Walter, Hongtao Yu, Brian Dale Ross and Alnawaz Rehemtulla Chandan Kumar-Sinha, Mukesh Kumar Nyati, Dipankar Ray, NilsJoseph Dosch, Marcian E. Van Dort, Sooryanarayana Varambally,

Chaudhry,Pitchiaya, Katerina Chekhovskiy, Areeb Chator, Nauman Shyam Nyati, Katrina Schinske-Sebolt, Sethuramasundaramsignaling

βThe kinase activity of the Ser/Thr kinase BUB1 promotes TGF-

This information is current as of January 6, 2015. The following resources related to this article are available online at http://stke.sciencemag.org.

Article Toolshttp://stke.sciencemag.org/content/8/358/ra1article tools: Visit the online version of this article to access the personalization and

MaterialsSupplemental

http://stke.sciencemag.org/content/suppl/2015/01/02/8.358.ra1.DC1.html"Supplementary Materials"

Related Content

http://stke.sciencemag.org/cgi/cm/stkecm;CMP_9876http://stke.sciencemag.org/cgi/cm/stkecm;CMN_9915http://stke.sciencemag.org/content/sigtrans/1/46/eg8.full.htmlhttp://stke.sciencemag.org/content/sigtrans/7/345/eg3.full.htmlhttp://stke.sciencemag.org/content/sigtrans/7/335/rs5.full.htmlhttp://stke.sciencemag.org/content/sigtrans/7/345/ra91.full.htmlhttp://stke.sciencemag.org/content/sigtrans/3/119/tr4.full.htmlhttp://stke.sciencemag.org/content/sigtrans/7/344/re8.full.html

's sites:ScienceThe editors suggest related resources on

Referenceshttp://stke.sciencemag.org/content/8/358/ra1#BIBLThis article cites 61 articles, 24 of which you can access for free at:

Glossaryhttp://stke.sciencemag.org/cgi/glossarylookupLook up definitions for abbreviations and terms found in this article:

Permissionshttp://www.sciencemag.org/about/permissions.dtlObtain information about reproducing this article:

reserved. DC 20005. Copyright 2015 by the American Association for the Advancement of Science; all rightsAmerican Association for the Advancement of Science, 1200 New York Avenue, NW, Washington,

(ISSN 1937-9145) is published weekly, except the last December, by theScience Signaling

on January 6, 2015http://stke.sciencem

ag.org/D

ownloaded from

www.sciencesignaling.org/cgi/content/full/8/358/ra1/DC1

Supplementary Materials for

The kinase activity of the Ser/Thr kinase BUB1 promotes TGF-β signaling

Shyam Nyati, Katrina Schinske-Sebolt, Sethuramasundaram Pitchiaya, Katerina

Chekhovskiy, Areeb Chator, Nauman Chaudhry, Joseph Dosch, Marcian E. Van Dort, Sooryanarayana Varambally, Chandan Kumar-Sinha, Mukesh Kumar Nyati,

Dipankar Ray, Nils G. Walter, Hongtao Yu, Brian Dale Ross, Alnawaz Rehemtulla*

*Corresponding author. E-mail: alnawaz@med.umich.edu

Published 6 January 2015, Sci. Signal. 8, ra1 (2015) DOI: 10.1126/scisignal.2005379

This PDF file includes:

Fig. S1. siRNA screening and pathway analysis. Fig. S2. BUB1 mediates TGF-β–dependent SMAD2/3 phosphorylation. Fig. S3. Depletion of BUB1 leads to reduced SMAD3 recruitment to TGFBRI and TGF-β–dependent accumulation of SMAD2 in the nucleus. Fig. S4. Depletion of BUB1 attenuates TGF-β–dependent SMAD2/3-SMAD4 complex formation in A549 cells. Fig. S5. Depletion of BUB1 inhibits TGF-β–mediated EMT in A549 and NCI-H358 cells. Fig. S6. Depletion of BUB1 inhibits TGF-β–mediated migration and invasion of A549 cells. Fig. S7. BUB1 kinase activity mediates TGF-β–dependent phosphorylation and nuclear activity of R-SMAD. Fig. S8. TGFBRI and BUB1 colocalization by TIRF microscopy at 1 and 24 hours and coimmunoprecipitation of His-TGFBRI and Myc-BUB1. Fig. S9. BUB1 kinase activity mediates TGFBRI-TGFBRII interaction. Fig. S10. TGFBRI is not a direct substrate of BUB1 kinase activity. Fig. S11. Wild-type Myc-BUB1 interacts with FL-SMAD2 in HEK293T cells. Fig. S12. SMAD3 is not a direct substrate of BUB1 kinase activity. Legends for tables S1 and S2 Table S3. Pathway impact analysis based on the fold induction of the BTR reporter.

Other Supplementary Material for this manuscript includes the following: (available at www.sciencesignaling.org/cgi/content/full/8/358/ra1/DC1)

Table S1 (Microsoft Excel format). Human kinome siRNA screen in A549-BTR and MDA-231-1833–BTR cells. Table S2 (Microsoft Excel format). Hits obtained in A549-BTR and MDA-231-1833–BTR human kinome screen.

Figure S1. siRNA screening and pathway analysis. (A) Schematic representation of siRNA screening strategy. siGENOME human kinome siRNA library was transfected in reporter expressing A549-BTR and MDA-231-1833-BTR cells (in triplicates) using Dharmafect1 reagent and a Beckman Coulter Biomek liquid handling system. Cells were incubated for 60 hours after transfection, serum-starved overnight, and TGF-β (10 ng/mL) was added for 1 hour, after which reporter activity was assessed upon addition of D-Luciferin (50 µg/ml) using an InVision plate reader. (B and C) Two-way evidence plots of high and low stringency hits generated by Pathway-Guide (top) and perturbation accumulation analysis of a selected pathway, the oocyte meiosis pathway (middle and bottom) in A549-BTR cells (B) and MDA-231-1833-BTR cells (C). Two-way evidence plots show the hypergeometric (over-representation, PNDE) vs. perturbation p-values (PPERT) for all the pathways. Pathways above the oblique red line are significant at 5% after Bonferroni correction; those above the oblique blue line are significant at 5% after FDR correction. The vertical and horizontal thresholds represent the same corrections for the two types of evidence considered individually. Oocyte meiosis pathway analysis (middle) shows the highest total perturbation accumulation (pA for BUB1, red dot). X-axis is Log2 FC (fold

change; BTR fold activation), and the Y-axis is perturbation accumulation (Acc) for a given gene in the pathway. Genes (green dots) on the gray line (along the Y-axis are) are not input genes, therefore they show only perturbation accumulation but no fold change values. To better understand the perturbation p-value of this selected pathway, its total accumulation is shown on the distribution of the total accumulations under the null hypothesis (bottom).

Figure S2. BUB1 mediates TGF-β–dependent SMAD2/3 phosphorylation. Lung (A549), and breast (MDA-231-1833, MCF7) cancer as well as non-tumoregenic lines (MRC5, MCF10A, Het1A) were transfected with scrambled control (NSS, negative control), TGFΒRI (positive control), or BUB1 siRNA and treated with TGF-β (10 ng/mL) for an hour. Extracts were prepared and analyzed for phosphorylated SMAD2 and SMAD2 abundance by immunoblot analysis. Data are means ± SD from at least three biological replicates (TGF-β-treated samples only).

Figure S3. Depletion of BUB1 leads to reduced SMAD3 recruitment to TGFBRI and TGF-β–dependent accumulation of SMAD2 in the nucleus. (A to B). Western blot quantitation of data in Fig. 2A (HEK293T) and Fig. 2B (A549). Data are means ± SD from 3 biological replicates. (C) Western blot of FL-SMAD3 immunoprecipitated with TGFBRI antibody in A549 cells transfected with FL-SMAD3, His-TGFBRI and siRNA to TGFBRI or BUB1, and treated with TGF-β (10 ng/mL, 1 hour). Blots are representative of 2 independent experiments. (D) Immunofluorescence for SMAD2 in A549 and NCI-H358 cells transfected for 60 hours with control or TGFBRI- or BUB1-specific siRNA. Cells were serum-starved overnight and treated with TGF-β or vehicle (mock) for 1 hour. DAPI counter-stained the nuclei (blue). Scale bar, 25 µm. Images are representative of 3 independent experiments.

Figure S4. Depletion of BUB1 attenuates TGF-β–dependent SMAD2/3-SMAD4 complex formation in A549 cells. Quantitative analysis of immunoblot results in Fig. 2C. Data are means ± S.E.M. from 3 biological replicates.

Figure S5. Depletion of BUB1 inhibits TGF-β–mediated EMT in A549 and NCI-H358 cells. Light microscopy to assess morphology of A549 (A and B) and NCI-H358 (D and E) cells transfected with control siRNA (NSS), TGFBRI siRNA, or BUB1 siRNA and either mock-treated or treated with TGF-β (10 ng/mL) for

72 hours. Images of 3 random fields for each treatment were taken at 40X objective. Data are means ± S.E.M. from 3 biological replicates. **P<0.001, two-sided Student’s t-test. Scale bars, 0.25 mm (A), 0.2 mm (D). (C) Representative Western blots for E-Cadherin and N-Cadherin in lysates from A549 cells transfected and treated as in (A and B). Blots are representative of 3 biological replicates.

Figure S6. Depletion of BUB1 inhibits TGF-β–mediated migration and invasion of A549 cells. (A to D) Migration assays using 0.8 µM filter wells (A and B) and invasion assays using Matrigel-coated wells (C and D) in A549 cells transfected as indicated with control (NSS) or targeted siRNA in the absence or presence of TGF-β (10 ng/mL) for 72 hours. Cell migration and invasion were counted from two random fields for each treatment. Data are means ± S.E. from 3 biological replicates. *P<0.001, **P<0.01, two-sided Student’s t-test. Images are representative. Scale bars, 100 µM.

Figure S7. BUB1 kinase activity mediates TGF-β–dependent phosphorylation and nuclear activity of R-SMAD. (A) Quantitation of TGF-β-treated western blot data in Fig. 3A. Data are means ± S.E. from 2 biological replicates. (B to D) Quantitative analysis of immunoblot results in Fig. 3 (C to E). In A549 (B), NCI-H358 (C), and MDA-231-1833 (D) cells, pSMAD2/3 abundance and 2OH-BNPP1 concentration showed a strong negative correlation (r) and a strong goodness of fit (R2). Data are means ± S.D. of 3 biological replicates for each cell lines. (E and F) Immunoblots of lysates from Het1A (E) and MRC5 (F) cells treated with vehicle, vehicle and TGF-β (10 ng/mL), or TGF-β and the indicated concentration of 2OH-BNPP1 for 1 hour. (G and H) Immunofluorescence analysis of SMAD2 in A549 (G) and NCI-H358 (H) cells treated with vehicle (mock), vehicle and TGF-β (10 ng/mL), or TGF-β and either SB-431542 (10 µM) or 2OH-BNPP1 (10 µM) for 1 hour. Nuclei were counter-stained with DAPI. Data are representative of 3 biological replicates. Scale bar, 25µm. (I and J) Relative firefly luciferase activity (normalized to Gaussia luciferase) in MDA-231-1833 (I) and HeLa (J) cells transiently transfected with the SBE4-Luc reporter plasmid and a GLuc plasmid and treated as in (G and H) for 24 hours. Data are means ± S.E.M. of 3 biological replicates. **P<0.001, two-sided Student’s t-test.

Figure S8. TGFBRI and BUB1 colocalization by TIRF microscopy at 1 and 24 hours and coimmonoprecipitation of His-TGFBRI and Myc-BUB1. (A) Pseudocolored, background-corrected, TIRF images of A549 cells that were treated with a mock reagent (top panel) or TGF-β (10 ng/mL, bottom panel) for 1 or 24 hours and immunostained for BUB1 (green) and TGFBRI (red). Scale bar, 10 µm. Insets (2.66x2.66µm2) re magnified images of the boxed regions. Line scan across BUB1 (green) and TGFBRI (red) particles within the inset are shown at the right. Data are representative of 3 experiments. (B and C) Quantitation of immunoblotting represented in Fig. 4D: coimmunoprecipitation of Myc-BUB1 with TGFBRI (B), or abundance of phosphorylated SMAD2 or phosphorylated SMAD3 (C). Data are means ± S.E. of 3 experiments. (D) Cell cycle analysis by propidium iodide staining in A549 cells transfected with control siRNA (NSS), TGFBRI siRNA, or BUB1 siRNA for 72 hours. Combined data from 3 biological replicates is shown.

Figure S9. BUB1 kinase activity mediates TGFBRI-TGFBRII interaction. Quantitation of Western blotting data in Fig. 4G. Data are means ± S.E. from 3 experiments.

Figure S10. TGFBRI is not a direct substrate of BUB1 kinase activity. In vitro kinase assay containing varying combinations of insect cell-purified full-length wild-type (WT) or kinase-deficient (KD) BUB1 and full-length wild-type or kinase-deficient (KD) TGFBRI. H2A was used as a positive control. Data are representative of 3 experiments.

Figure S11. Wild-type Myc-BUB1 interacts with FL-SMAD2 in HEK293T cells. Immunoprecipitation for Myc or BUB1 from lysates of HEK293T cells transfected with wild-type Myc-BUB1 and Flag-tagged (FL) SMAD2. Input lysates were probed with control antibodies. Blots are representative of 2 experiments.

Figure S12. SMAD3 is not a direct substrate of BUB1 kinase activity. In vitro kinase assay containing varying combinations of insect cell-purified full-length wild-type BUB1 or TGFBRI and E. coli–purified GST-tagged SMAD3. Data are representative of 3 experiments.

Table S1. Human kinome siRNA screen in A549-BTR and MDA-231-1833–BTR cells. (Table is supplied as an Excel file in additional supplementary files online). BTR-fold change over non-targeting scrambled control (NSS) siRNA transfected cells is calculated for each replicate for each cell-line. Average of three replicates is calculated and the data is sorted based on the average reporter fold induction over control. Plate number for siRNA to specific kinase is provided in the first column. The data is deposited at PubChem (https://pubchem.ncbi.nlm.nih.gov/assay/assay.cgi?aid=1117269), accession ID 1117269. Table S2. Hits obtained in A549-BTR and MDA-231-1833–BTR human kinome screen. (Table is supplied as an Excel file in additional supplementary files online). Hits obtained in A549-BTR (green) and MDA-231-1833-BTR (red). High stringency hits in experiment-wise analysis are shown in bold while low-stringency hits are in normal text. A reference for known association of a kinase with TGF-β signaling pathway is provided.

Table S3. Pathway impact analysis based on the fold induction of the BTR reporter. (Table is on the next page, below). Low and high-stringency hits (listed in table S2) were subjected to Pathway Impact Analysis. Rank, name, pathway ID, pathway size (pSize), number of differentially expressed genes (NDE, genes subjected to analysis), total perturbation accumulation (tA), p-values for over representation of differentially expressed genes in a given pathway (PNDE,), the degree of perturbation (PPERT), overall p-value (pG, combining PNDE and PPERT), overall FDR corrected p-value (pGFDR), overall FWER corrected p-value (pG FWER) are given.

Table S3. Pathway impact analysis based on the fold induction of the BTR reporter.

Rank Pathway (A549-BTR)

Pathway ID

pSize NDE tA pNDE pPERT pG pG FDR pG FWER

1 Insulin signaling pathway

4910 138 7 -4E+01 4E-12 0.35 4.06E-11

2.97E-09 2.97E-09

2 mTOR signaling pathway

4150 64 4 -2E+01 1E-07 0.085 1.65E-07

6.00E-06 1.20E-05

3 Adipocytokine signaling pathwa

4920 69 4 -6E+00 1E-07 0.647 1.50E-06

2.99E-05 1.09E-04

4 ErbB signaling pathway

4012 88 4 -4E+01 4E-07 0.266 1.64E-06

2.99E-05 1.20E-04

5 MAPK signaling pathway

4010 259 5 -2E+01 8E-07 0.178 2.29E-06

3.34E-05 1.67E-04

6 Progesterone-mediated oocyte maturation

4914 86 4 4E-01 3E-07 0.939 4.91E-06

5.97E-05 3.58E-04

7 Neurotrophin signaling pathway

4722 120 4 -9E+00 1E-06 0.51 9.69E-06

1.01E-04 7.07E-04

8 Type II diabetes mellitus

4930 48 3 -8E+00 5E-06 0.709 4.38E-05

4.00E-04 3.20E-03

9 Toxoplasmosis 5145 118 3 3E+01 7E-05 0.068 6.12E-05

4.96E-04 4.47E-03

10 Influenza A 5164 174 3 3E+01 2E-04 0.032 8.81E-05

6.43E-04 6.43E-03

11 Fc gamma R-mediated phagocytos

4666 95 3 -9E+00 4E-05 0.341 1.49E-04

9.89E-04 1.09E-02

12 Phosphatidylinositol signaling

4070 81 3 8E-17 2E-05 0.991 2.56E-04

1.56E-03 1.87E-02

13 Tuberculosis 5152 179 3 4E+01 2E-04 0.111 2.98E-04

1.67E-03 2.17E-02

14 HIF-1 signaling pathway

4066 110 3 4E+00 5E-05 0.658 4.06E-04

2.12E-03 2.96E-02

15 PI3K-Akt signaling pathway

4151 338 4 2E+01 7E-05 0.632 5.08E-04

2.47E-03 3.71E-02

16 Chemokine signaling pathway

4062 189 3 2E+01 3E-04 0.417 1.14E-03

5.22E-03 8.35E-02

17 Focal adhesion 4510 204 3 -2E+01 3E-04 0.42 1.41E-03

5.87E-03 1.03E-01

18 Bacterial invasion of epitheli

5100 70 2 4E+01 1E-03 0.146 1.45E-03

5.87E-03 1.06E-01

19 Circadian rhythm 4710 30 2 8E-01 2E-04 0.969 1.72E-03

6.63E-03 1.26E-01

20 Cholinergic synapse 4725 112 2 6E+01 3E-03 0.076 1.85E-03

6.76E-03 1.35E-01

31 Oocyte meiosis 4114 112 2 3E+01 3E-03 0.428

8.55E-03 2.01E-02 6.24E-01

32 TGF-beta signaling pathway

4350 83 2 -6E-14 1E-03 0.836

9.15E-03 2.09E-02 6.68E-01

Rank Pathway (MDA-

231-1833 BTR) Pathway

ID pSize NDE tA pNDE pPERT pG pG FDR pG FWER

1 TGF-beta signaling pathway 4350 83 3 -7E+01 2E-05 0.309

9.38E-05 6.76E-03 6.76E-03

2 ErbB signaling pathway 4012 88 3 3E+01 3E-05 0.773

2.56E-04 9.21E-03 1.84E-02

3 mTOR signaling pathway 4150 64 2 -4E+01 8E-04 0.063

5.77E-04 1.38E-02 4.15E-02

4 Phototransduction 4744 29 1 -1E+02 2E-02 0.005 9.83E-

04 1.77E-02 7.08E-02

5 Phosphatidylinositol signaling 4070 81 2 1E+01 1E-03 0.101

1.35E-03 1.80E-02 9.71E-02

6 Chemokine signaling pathway 4062 189 3 -5E+01 3E-04 0.565

1.50E-03 1.80E-02 1.08E-01

7 Leishmaniasis 5140 72 2 3E+01 1E-03 0.256 2.52E-

03 2.42E-02 1.81E-01

8 Acute myeloid leukemia 5221 57 2 -1E+01 7E-04 0.439

2.69E-03 2.42E-02 1.94E-01

9

Progesterone-mediated oocyte maturation 4914 86 2 3E+01 2E-03 0.351

4.55E-03 3.47E-02 3.27E-01

10 Toxoplasmosis 5145 118 2 4E+01 3E-03 0.211 5.03E-

03 3.47E-02 3.62E-01

11 HIF-1 signaling pathway 4066 110 2 2E+01 2E-03 0.258

5.32E-03 3.47E-02 3.83E-01

12 Influenza A 5164 174 2 4E+01 6E-03 0.116 5.78E-

03 3.47E-02 4.16E-01

13 Herpes simplex infection 5168 184 2 3E+01 7E-03 0.118

6.45E-03 3.57E-02 4.64E-01

14 Oocyte meiosis 4114 112 2 8E+01 3E-03 0.348 7.13E-

03 3.67E-02 5.14E-01

15 Cholinergic synapse 4725 112 2 6E+01 3E-03 0.408 8.20E-

03 3.93E-02 5.90E-01

16 Dorso-ventral axis formation 4320 24 1 -1E+01 2E-02 0.072

8.90E-03 4.00E-02 6.41E-01

17 PI3K-Akt signaling pathway 4151 338 3 2E+00 1E-03 0.985

1.09E-02 4.61E-02 7.84E-01

18 Insulin signaling pathway 4910 138 2 3E+01 4E-03 0.437

1.24E-02 4.84E-02 8.93E-01

19

Fc gamma R-mediated phagocytos 4666 95 2 1E+00 2E-03 0.939

1.28E-02 4.84E-02 9.19E-01

20 Tuberculosis 5152 179 2 5E+01 6E-03 0.306 1.41E-

02 5.08E-02 1.00E+00