LEO1 Is Regulated by PRL-3 and Mediates Its Oncogenic ... · Molecular and Cellular Pathobiology...

Molecular and Cellular Pathobiology

LEO1 Is Regulated by PRL-3 and Mediates Its OncogenicProperties in Acute Myelogenous Leukemia

Phyllis S.Y. Chong1, Jianbiao Zhou1, Lip-Lee Cheong2, Shaw-Cheng Liu1, Jingru Qian3,Tiannan Guo3, Siu Kwan Sze3, Qi Zeng4, and Wee Joo Chng1,2,5

AbstractPRL-3, an oncogenic dual-specificity phosphatase, is overexpressed in 50% of acute myelogenous leukemia

(AML) and associated with poor survival. We found that stable expression of PRL-3 confers cytokineindependence and growth advantage of AML cells. However, how PRL-3 mediates these functions in AML isnot known. To comprehensively screen for PRL3-regulated proteins in AML, we performed SILAC-basedquantitative proteomics analysis and discovered 398 significantly perturbed proteins after PRL-3 over-expression. We show that Leo1, a component of RNA polymerase II–associated factor (PAF) complex, is anovel and important mediator of PRL-3 oncogenic activities in AML. We described a novel mechanism whereelevated PRL-3 protein increases JMJD2C histone demethylase occupancy on Leo1 promoter, therebyreducing the H3K9me3 repressive signals and promoting Leo1 gene expression. Furthermore, PRL-3 andLeo1 levels were positively associated in AML patient samples (N¼ 24; P < 0.01). On the other hand, inhibitionof Leo1 reverses PRL-3 oncogenic phenotypes in AML. Loss of Leo1 leads to destabilization of the PAFcomplex and downregulation of SOX2 and SOX4, potent oncogenes in myeloid transformation. In conclusion,we identify an important and novel mechanism by which PRL-3 mediates its oncogenic function in AML.Cancer Res; 74(11); 3043–53. �2014 AACR.

IntroductionPRL-3, encoded by the PTP4A3 gene, is a small dual-

specificity phosphatase characterized by the conservedC(X5)R catalytic domain, and a unique C-terminal prenyla-tion domain essential for its proper subcellular localization(1, 2). PRL-3 has been shown to promote cellular processes,such as cell motility, invasion, cell growth, and survival,through various mechanisms (2–7). PRL-3 was first linked tocancer when it was consistently found at elevated levels incolorectal cancer metastases, but at much lower levels inmatched early-staged tumor and normal colorectal epithe-lium (8). Since then, elevated expression of PRL-3 has beenimplicated in the progression and metastasis of an array ofcancer types, including gastric, ovarian, cervical, lung, liver,

and breast, and PRL-3 protein was overexpressed in anaverage of 22.3% of 1,008 human carcinoma samples exam-ined using immunohistochemistry (9, 10). Together with thefact that it has a highly restricted basal pattern of expressionin adult tissues (11), PRL-3 is deemed as an attractivetherapeutic target that spares normal tissues. The potentialof this target has been demonstrated using PRL-3–specificantibodies in an in vivo model (12).

In recent years, accumulating evidence suggests that PRL-3is also a novel therapeutic target and biomarker in leukemia(13, 14). We were the first to report that elevated PRL-3 proteinexpression occurs in about 47% of human acute myelogenousleukemia (AML) cases while absent from normal myeloid cellsin bone marrow (13). In addition, a large-scale gene expressionprofiling study of 454 primary AML samples demonstratesthat high PRL-3 levels is an independent negative prognosticfactor in AML, both for overall survival and event-free survival(14). These reports collectively suggest that PRL-3 may be ofbiologic and clinical relevance in AML and warrants furtherinvestigation.

In the present study, we created a TF-1 AML cell line stablyoverexpressing PRL-3 (TF1-hPRL3) as a model to study thebiologic relevance of PRL-3 in AML, and we use a quantitativeproteomic strategy to profile on a global scale changes inprotein expression induced by PRL-3 in AML. We found thatPRL-3 has pro-oncogenic properties in AML, and Leo1, acomponent of the humanRNApolymerase II–associated factor(PAF) complex, is one of the most differentially expressedproteins induced by PRL-3. Further, we showed that PRL-3

Authors' Affiliations: 1Cancer Science Institute of Singapore; 2Depart-ment of Medicine, Yong Loo Lin School of Medicine, National University ofSingapore; 3Department of Biological Sciences, Nanyang TechnologicalUniversity; 4Institute of Molecular and Cell Biology (A�STAR); and 5Depart-ment of Haematology-Oncology, National University Cancer Institute ofSingapore, National University Health System, Singapore

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

P.S.Y. Chong and J. Zhou contributed equally to this work.

Corresponding Author: Wee-Joo Chng, National University Health Sys-tem, 1EKentRidgeRoad,NUHSTowerBlock, Level 10,Singapore 119228.Phone: 65-6772-4612; Fax: 65-6777-5545; Email: [email protected]

doi: 10.1158/0008-5472.CAN-13-2321

�2014 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 3043

on November 17, 2020. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst March 31, 2014; DOI: 10.1158/0008-5472.CAN-13-2321

http://cancerres.aacrjournals.org/

upregulates Leo1 through a novel epigenetic mechanism. Onthe other hand, knockdown of Leo1 significantly diminishesthe oncogenic effects of PRL-3. Our current work implicates apro-oncogenic role of PRL-3 in AML, and reveals Leo1 as anovel downstream molecule required for PRL-3 oncogenicfunction in leukemia.

Materials and MethodsCell culture

HEK293T cells were cultured in Dulbecco's Modified EagleMediumwith 10% fetal calf serum (FCS). TF1-derived cell lineswere cultured in RPMI 1640medium containing 10%FCS (R10)supplemented with 5 ng/mL human interleukin (IL)-3 (Milte-nyi Biotec). Molm-14, HEL, and HL-60 were cultured in R10.Human CD34þ cells were grown in StemSpan SFEM IImediumsupplemented with StemSpan CC100 cytokine cocktail (Stem-Cell Technologies). Primary AML cells were grown in sameconditions, with the addition of granulocyte macrophagecolony-stimulating factor (20 ng/mL). Cell lines were obtainedfrom American Type Culture Collection and authenticated.Plasmid details are available in Supplementary Methods.

SILAC-based mass spectrometryTF-1 was cultured in "light" stable isotope labeling by amino

acids in cell culture (SILAC) medium containing normal lysineand arginine amino acids, whereas the TF1-hPRL3 was grownin "heavy" SILAC medium with stable isotope-labeled 13C6

lysine (þ6-Da shift) and 13C615N4 arginine (þ10-Da shift;

Thermo Scientific). The cellular lysates were combined andproteolytically digested by trypsin followed by tandem massspectrometry (MS) identification. Peptides were subjected totarget-decoy database search strategy with a false discoveryrate (FDR) of less than 1% (<1%). Differential protein expres-sion was quantified from the relative intensity ratios in the MSspectra between the "heavy" and "light" states.

Xenograft mode modelsSix-week-old female nonobese diabetic/severe combined

immunodeficient (NOD/SCID) mice were provided by Dr.Chan Shing-Leng (CSI Singapore, NUS). Exponentially growingTF1-pEGFP and TF1-hPRL3 cells (5 � 106) were subcutane-ously injected into loose skin between the shoulder blades andthe left and right front leg of NOD/SCID-recipient mice (3micetotal), respectively. The length (L) and width (W) of the tumorwere measured with callipers, and tumor volume (TV) wascalculated as TV ¼ (L � W2)/2. The protocol is reviewed andapproved by Institutional Animal Care and Use Committee incompliance to the guidelines on the care and use of animals forscientific purpose.

Transfection and lentiviral-mediated shRNA deliveryHigh-titre lentiviruses were produced in HEK293T cells by

transfection using JetPrime (Polyplus) with packaging vectorspsPAX2 and pMD2.G togetherwith the respective short hairpinRNAs (shRNA) constructs. Virus was harvested, pooled, andconcentrated by centrifugation (Amicon). Infection of leuke-mia cells was done by the spin-infection method.

RNA extraction and quantitative reverse transcriptionPCR

Detailed protocol is available in Supplementary Methods.

Western blottingCells were counted and lysed in radioimmunoprecipita-

tion assay buffer. Anti-PRL3 (clone 318) was a kind gift fromDr. Zeng Qi [Institute of Molecular and Cell Biology, Agencyfor Science, Technology, and Research (A*STAR), Singa-pore]; anti-Leo1, anti-Paf1, and anti-Ctr9 antibodies werefrom Bethyl Laboratories; anti-actin, anti-hnRNPE1, anti-HSP90, anti-JMJD2C, anti-GFP, and anti-c-Myc (9E10) anti-bodies were from Santa Cruz Biotechnology; anti-Stathmin, anti-HDAC2 and anti-H3, anti-H3K9me2, anti-H3K27me2, anti-H3K27me3, anti-H3K4me1, anti-H3K4me2,anti-H3K4me3, and anti-H3K79me2 antibodies were fromCell Signaling Technology; anti-H3K9me3 antibody wasfrom Active Motif.

Chromatin immunoprecipitationChromatin immunoprecipitation (ChIP) assays were per-

formed using the respective antibodies according to the man-ufacturer's instructions (Thermo Scientific). Quantitative PCR(qPCR) was performed using the eluted DNA (sample) and 1%input, with primers spanning Leo1 promoter region withrespect to Leo1 transcriptional start site. The percentage ofinput was calculated as described previously (15). Primersequences used for ChIP–qPCR are available in SupplementaryTable S3.

Luciferase assayDetailed protocol is available in Supplementary Methods.

Cell proliferation, apoptosis, and CFADetailed protocol is available in Supplementary Methods.

ResultsPRL-3 promotes cytokine-independent growth, colony-forming capacity of AML cells in vitro andtumorigenecity in vivo

To assess the roles of PRL-3 in pathogenesis of AML, wedeveloped a pair of stable, isogenic cell lines, TF1-pEGFP andTF1-hPRL3 by transfecting pEGFP (vector control) andpEGFP-hPRL-3 vectors into TF-1 cells, respectively, followedby G418 selection and fluorescence-activated cell sorting(Fig. 1A). TF-1 is a cytokine-dependent AML cell line. Quan-titative reverse transcription PCR (qRT-PCR) and Westernblot validated the overexpression of PRL-3 on both mRNA andprotein levels in the TF1-hPRL3 cells relative to TF1-pEGFPcells (Fig. 1B). In the absence of cytokine (human IL-3), themajority of TF1-pEGFP cells became apoptotic after 72 hours.In contrast, most of TF1-hPRL3 cells were viable (Fig. 1C). TF1-hPRL3 cells not only resisted cytokine deprivation-inducedapoptosis, but also proliferated well without additional cyto-kines (Fig. 1C). Furthermore, methylcellulose assay showedonly TF1-hPRL3 cells, but not TF1-pEGFP cells, formed colo-nies (Fig. 1D). To determine the tumorigenic role of PRL-3 in

Chong et al.

Cancer Res; 74(11) June 1, 2014 Cancer Research3044




vivo, we used the pair of TF1 cells in a subcutaneous mousexenograft model. Strikingly, only TF1-hPRL3 cells, but not TF1-pEGFP cells, formed tumor mass in immunodeficient mice(Fig. 1E). The average TF1-hPRL3 tumor volume was 813 mm3

at 4 weeks after cell inoculation (Fig. 1E). Overall, these datasuggest that PRL-3 confers important oncogenic function inAML cells in vitro and in vivo.

SILAC quantitative proteomic profiling of PRL-3overexpressed cellTo understand downstream pathways induced by over-

expressing PRL-3, we performed a quantitative proteomic

study by combining metabolic labeling using SILAC andhigh-resolution MS approach (Supplementary Fig. S1A). Weidentified 398 proteins whose expression was significantlyperturbed by PRL-3 overexpression, with 331 proteins upre-gulated and 67 proteins downregulated (>1.5-fold; Supple-mentary Fig. S1B).

To gain a functional understanding of the dataset, wesubjected the differentially expressed proteins (up- or down-regulated) to gene ontology analysis. Upregulated proteins areenriched for processes such as nucleic acid metabolism(19.7%), protein metabolism (15.1%), and pre-mRNA proces-sing (5.7%), which are novel biologic processes associated with

Bright field GFP

TF1-pEGFP

TF1-hPRL3

14

12

10

8

6

4

2

0

**

**

TF1-EGFP

TF1-EGFP

TF1-EGFP

TF1-hPRL3

TF1-hPRL3

TF1-pEGFP TF1-hPRL3

TF1-hPRL3

Rela

tive e

xp

ressio

n

TF1-EGFP

TF1-hPRL3

PRL-3

b-Actin

120

100

80

60

40

20

0

Via

ble

cells (

%)

Day 0 Day 2 Day 4 Day 6

60

50

40

30

20

10

0

Cell

nu

mb

er

(×10

5)

Without additional cytokine

250

200

150

100

50

0

Nu

mb

er

of

co

lon

ies

(/w

ell)

1,400

1,200

1,000

800

600

400

200

0

Tu

mo

r vo

lum

e (

mm

3)

**

**A D

B E

C

Figure 1. Establishment of stable PRL-3 cell line. A, images of TF1-pEGFP and TF1-hPRL3 cells in bright field and GFP channel under an invertmicroscopy. B, qRT-PCR and Western blot analysis of PRL-3 mRNA and protein expression in TF1-pEGFP and TF1-hPRL3 cells. b-Actin was used as aloading control in the protein analysis. C, after IL-3 starvation, TF1-pEGFP and TF1-hPRL3 cells were stained and assessed with Trypan BlueExclusion methods on a hemacytometer. The percentages of viable cells at 72 hours were calculated relative to before IL-3 withdrawal (0 hour). Toconstruct the cell growth curve, 20 � 105 TF1-hPRL3 cells were initially seeded in standard medium without additional human cytokines. Cells werecounted and reseeded in fresh medium every 2 days for up to 6 days. In both figures, three independent replicates were conducted (mean � SD). D,colony-forming assay of TF1-pEGFP and TF1-hPRL3. The experiments were duplicated and representative pictures are presented. E, mouse xenograftmodels of TF1-pEGFP and TF1-hPRL3. Five million TF1-pEGFP and TF1-hPRL3 cells were subcutaneously injected into loose skin between theshoulder blades and the left and right front leg of NOD/SCID recipient mice, respectively. At 4 weeks after injection, tumors were harvested andmeasurement of tumor volume was taken. The tumor volume is shown as the mean � SD (mm3). Arrows, the TF1-hPRL3 tumor mass at right andno tumor at left (TF1-pEGFP cell injected site) in one mouse. ��, P < 0.005.

LEO1 Mediates PRL-3 Oncogenic Function in AML

www.aacrjournals.org Cancer Res; 74(11) June 1, 2014 3045




PRL-3, as well as known involvement in cell-cycle regulation(7.1%) and cell motility (6%) as depicted in Supplementary Fig.S1C. Amongst the differentially regulated proteins, 45.5% ofthem are nuclear proteins, whereas 54.5% are cytoplasmicproteins (Supplementary Fig. S1D).

Validation of MS resultsExpression of several representative proteins, including

Leo1, stathmin, HSP-90, hnRNPE1, and HDAC2, was com-pared between TF1-pEGFP and TF1-hPRL3 cells (Fig. 2A),and concordance was observed between the Western blotand MS results. Unexpectedly, we did not detect PRL-3 inour liquid chromatography/tandem mass spectrometry(LC/MS-MS) result. This is likely due to the trypsin cleavagesites distribution in the PRL-3 that generates peptideseither too long or too short for detection by LC-MS/MS.Nonetheless, PRL-3 overexpression was confirmed at boththe mRNA (data not shown) and protein levels by Westernblot (Fig. 2A).

We further performed lentiviral-based delivery of PRL-3shRNA into two additional human AML cells, Molm-14 andHEL, which expresses high endogenous level of PRL-3.The knockdown of PRL-3 reduced the levels of Leo1,stathmin, HSP90, and hnRNPE1 (Fig. 2B). An exception wasHDAC2, where although MS identifies it as one of the

upregulated proteins by PRL-3, Western blot was not sen-sitive enough to detect the differences (Fig. 2B). Theseexperiments confirmed the general validity of our proteo-mic analysis.

Leo1 is upregulated by PRL-3Leo1, a component of the human PAF complex, was the

top candidate from the SILAC LC-MS/MS analysis with themost significant increase in protein abundance. We there-fore decided to focus on this protein in our subsequentexperiments. To confirm that Leo1 is a downstream target ofPRL-3, we introduced a Tet-On myc-tagged PRL-3 constructinto 293T cells. Forty-eight hours after transfection, 293Tcells were exposed to increasing concentrations of thetetracycline derivative, doxycycline, to induce PRL-3 expres-sion. PRL-3 protein accumulated in a dose-dependent man-ner, and a corresponding increase in Leo1 protein levels wasobserved (Fig. 2C) and quantified (Supplementary Fig. S2).To complement the overexpression studies, we also per-formed knockdown by transfecting PRL-3 siRNA into 293Tcells. The degree of PRL-3 inhibition correlated to the dose ofsiRNA added, and Leo1 was downregulated (Fig. 2D). Col-lectively, these observations corroborated the MS resultsshowing that Leo1 expression is positively correlated withPRL-3 levels.

TF1TF1-P

RL3

Molm-14

Molm14-P

RL3 KD

HEL-PRL3 K

D

HEL

Leo1

Leo1

Leo1

Leo1

Stathmin Stathmin

HSP-90HSP-90

hnRNPE1 hnRNPE1

HDAC2 HDAC2

PRL-3

PRL-3

PRL-3

b-Actin

b-Actin b-Actin

Actin

Emp 0 2 20 (mg/mL Dox)

Myc (PRL3)

Mock

1.4 mg s

iRNA

2.8 mg s

iRNA

0.7 mg s

iRNA

0.3 mg s

iRNA

A B

C D

Figure 2. Western blot validation ofcandidate proteins identified byMS. A, expression of severalrepresentative upregulatedproteins is compared betweenTF-1 and TF1-hPRL3. B,knockdown of endogenous PRL-3in AML cell linesMolm-14 and HEL.The same panel of proteins isexamined byWestern blot. C, 293Tcells were stably transfected withan inducible Tet-on myc-taggedPRL-3 construct, and increasingconcentration of doxycyclin wasused to induce PRL-3 expression.D, 293T cells were transfected withspecified doses of PRL-3 siRNA for48 hours before analysis withWestern blot.

Chong et al.





Abrogation of Leo1 reduces PRL3-mediatedproliferation and cytokine independence of AMLTo determine whether Leo1 contributes to PRL3-induced

leukemic growth, we used lentiviral-based delivery of shRNAsto deplete Leo1 expression and looked at growth properties.qRT-PCR and immunoblotting analysis demonstrated a 5-foldreduction in Leo1 mRNA levels and a marked reduction in theLeo1 protein levels upon treatment with Leo1 shRNA (Fig. 3A).Notably, reduction in Leo1 levels significantly impaired growthof TF1-hPRL3 cells to growth rates comparable with TF-1parental cells (Fig. 3B).To determine whether loss of Leo1 affects the survival of

TF1-hPRL3, we plated the cells in the absence of human IL-3 for 72 hours, followed by flow cytometric analysis for

apoptotic markers Annexin V-phycoerythrin and 7-ami-noactinomycin D (7-AAD). Although approximately 90%of TF1-hPRL3 population was able to survive withoutsupplementation of cytokine, Leo1 knockdown notablyincreased the apoptosis of TF1-hPRL3 induced by cytokinewithdrawal (Fig. 3C), indicating that Leo1 abrogationremoves the protective effect of PRL-3 toward cytokinedeprivation. Leo1 knockdown in PRL-3 cells resulted insignificantly fewer colonies, indicating a loss of clonogenicsurvival of these cells (Fig. 3D). These findings collectivelyindicate that Leo1 is an important mediator of the cellgrowth and survival of the TF1-hPRL3 cells, and loss of Leo1is largely sufficient to reverse these properties conferred byPRL-3 overexpression.

Rela

tive q

uan

tifi

cati

on

% A

po

pto

tic c

ells

Nu

mb

er

of

co

lon

ies (

/well)

Lu

min

escen

ce

Time (h)

1.4

1.2

1

0.8

0.6

0.4

0.2

0TF1-EGFP TF1-PRL3_Scram TF1-PRL3_Leo1

shRNA

Leo1

b-Actin

**

0 24 48 72 96

TF1-PRL3_Scram

TF1-EGFP

TF1-PRL3_Leo1

shRNA

10,000,000

9,000,000

8,000,000

7,000,000

5,000,000

6,000,000

4,000,000

3,000,000

1,000,000

2,000,000

0

100

90

80

70

60

50

40

30

20

10

0TF1-EGFP

TF1-EGFPTF1-PRL3_Scram

TF1-PRL3_Scram TF1-PRL3_Leo1TF1-PRL3_Leo1

shRNAshRNA

250

200

150

100

50

0

****

A B

C D

Figure 3. Effects of stable Leo1 knockdown in TF1-hPRL3 cells. A, knockdown of Leo1 in TF1-hPRL3 cells were analyzed with qRT-PCR and Westernblot. b-Actin was used as loading control. B, growth curve was determined by CellTitre-Glo luminescent assay. Cells were plated in 96-wellformat and grown for up to 4 days. Experiments were performed in triplicates. C, TF-1, PRL3-scrambled, and PRL3-Leo1 shRNA cells were seededin medium without human IL-3 for 48 hours. The graph is plotted with the percentages of apoptotic cells defined by Annexin V and 7-AAD positivity.D, TF-1, PRL3-scrambled, and PRL3-Leo1 shRNA cells were evenly suspended in HSC-colony-forming unit medium (StemMACS) and incubated for 10days for colonies to develop. ��, P < 0.005.






Leo1 and PRL-3 levels are significantly associated inAMLpatient samples

Because we observed a strong association between Leo1 andPRL-3 levels incell lines,wedeterminedwhetherLeo1 issimilarlyupregulated in human primary AML cells.Western blot analysisof PRL-3 and Leo1 protein expression in an independent cohortof 24 primary AML samples found that 50% of the patientsamples (Fig. 4A) were positive for PRL-3 protein. Of these, onlyone did not express Leo1. Therefore, a strong association existsbetweenPRL-3 andLeo1 expression (c2P value< 0.01). Leo1wasalsoexpressed infivesampleswithnoPRL-3expression, suggest-ing that other mechanisms may be responsible in these cases.Ourfindingsdemonstrated thata significantproportion (50%)ofAMLcases seemed tocoexpressPRL-3 andLeo1, suggesting thatLeo1 is a likely downstream target of PRL-3.

We also examined the basal expression levels of human PRL-3 and Leo1 in normal hematopoietic cell fractions, namelyCD34þ, T cell, B cell, and myeloid cell. Similarly, we found arelative correlation between PRL-3 and Leo1 levels. Further-more, we observed that PRL-3 and Leo1 are expressed at higherlevels in CD34þ and myeloid cells (Supplementary Fig. S3).

To validate the functional relevance of PRL-3 in humanAML,we downregulate PRL-3 in primary AML cells (Fig. 4B). PRL-3siRNA inhibited AML cell growth in a proliferation assay ascompared with transfection with control siRNA (Fig. 4C). Thisresult suggests that PRL-3 expression is also of biologic rele-vance in human AML and specific targeting of PRL-3 may havepossible therapeutic applications in AML.

PRL-3 affects Leo1 mRNA levels through histone H3lysine 9 (H3K9me3) demethylation

To examine whether PRL-3 regulates Leo1 at the transcrip-tional level, we perturbed PRL-3 with different shRNA con-

structs, whichwas accompanied by a concomitant reduction ofLeo1 transcripts levels (Fig. 5A). To confirm that PRL-3 reg-ulates Leo1 at the transcriptional level, we established a Leo1reporter construct containing a 703-bp Leo1 promoter frag-ment cloned upstream of the firefly luciferase gene (pGL3-Leo1). Transfection of the reporter construct, together withvector expressing wild-type PRL-3, increased the luciferaseactivity in a dose-dependent manner (Fig. 5B).

Given that PRL-3 is not a DNA-binding protein, thetranscriptional involvement of PRL-3 is likely to be indirect.A recent report implicated an epigenetic role of PRL-3 incolorectal cancer through regulating histone-modifyingenzymes involved in histone demethylation (16). To explorethe possibility that PRL-3 activates Leo1 through histonemethylation status in leukemic cells, we performed anunbiased screen of eight different permissive and repressivehistone modifications in the TF-1 and TF1-hPRL3 cells. Wefound that PRL-3 overexpression specifically leads to aglobal decrease in di- and trimethylation at histone H3lysine 9 (H3K9me2 and 3) residues (Fig. 5C). We furtherconfirmed this observation in another AML cell line, Molm-14 (Supplementary Fig. S4).

To determine whether increased Leo1 expression is associ-ated with local reductions in repressive histone methylation,we compared the basal levels of H3K9me2, H3K9me3, andH3K27me3 across Leo1 promoter region in the TF-1 and TF1-hPRL3 cells. In TF1-hPRL3 cells, a marked reduction inH3K9me3 and, to a lesser extent, in H3K9me2 was detectedthroughout the entire promoter region (P1–P4) as comparedwith the TF-1 cells (Fig. 5D). In contrast, H3K27me3 levelswere very low on the promoter (Supplementary Fig. S5). Thus,Leo1 promoter seems to be predominantly modulated by themethylation levels of H3K9.

Patient no#

Patient no#

Patient no#

1 2 3 4 5 6 7 8

9 10 11 12 13 14 15 16

17 18 19 20 21 22 23 24

Leo1

Leo1

Leo1

PRL-3

PRL-3

PRL-3

β-Actin

β-Actin

β-Actin

0 24 48 72

1.60

1.40

1.20

1.00

0.80

0.60

0.40

0.20

0.00

Re

lati

ve

gro

wth

CtsiRNA

siRNAPRL3

1.2

1

0.8

0.6

0.4

0.2

0Ct siRNA PRL3 siRNA

**

**

Re

lati

ve

qu

an

tifi

ca

tio

n

A B

C

Figure 4. Leo1 and PRL-3expression in human primary AMLcells. A, Western blot wasperformed to show the expressionlevels of Leo1 and PRL-3 in acohort of 24 primary AML samples.b-Actin was used as loadingcontrol. c2 test indicates highlysignificant association (P < 0.01).B, analysis of PRL-3 mRNAexpression in primary AML cellstransfected with either controlsiRNA or PRL-3 siRNA for 24hours. C, cell proliferation assay ofprimary AML cells transfected witheither control siRNA or PRL-3siRNA for 3 days. ��, P < 0.01

Chong et al.





PRL-3 mediates H3K9me3 demethylation on Leo1promoter through JMJD2C histone demethylaseNext, we sought to identify which histone demethylase is

involved in mediating this process. The members of theJMJD2 family, including the JMJD2A, JMJD2B, JMJD2C, andJMJD2D, are known to exhibit specific lysine-trimethyldemethylation activities (17). We systematically knockeddown these candidates and examined the impact onH3K9me3 levels and Leo1 expression. The JMJD2 shRNAsexhibited high specificity (Fig. 6A). We observed thatJMJD2C depletion in TF1-hPRL3 cells was able to increasethe global levels of H3K9me3, accompanied by a reduction inLeo1 (Fig. 6B), implying the specific involvement of JMJD2Cin regulating H3K9me3 related to levels of Leo1.To demonstrate that JMJD2C directly regulates H3K9me3

on the Leo1 promoter, we examined the levels of H3K9me3on region P2 on Leo1 promoter. Accordingly, knockdown ofJMJD2C reduced the JMJD2C signal on the Leo1 promoter

(Fig. 6C) and led to a 40% augmentation of the H3K9me3signal (Fig. 6D). However, knockdown of JMJD2A, JMJD2B,and JMJD2D did not modulate H3K9me3 levels on Leo1promoter (Supplementary Fig. S6A). Next, we examined howPRL-3 phosphatase regulates JMJD2C histone demethylase.JMJD2C protein expression did not increase in TF1-hPRL3cells (Supplementary Fig. S6B). We also examined the geneexpression levels of JMJD2A, JMJD2B, JMJD2C, and JMJD2Din TF1 and TF1-PRL3 cells, and found no significant changesbetween the isogenic cell lines (Supplementary Fig. S6C).Furthermore, PRL-3 and JMJD2C did not physically interact(Supplementary Fig. S6D). However, an increase in JMJD2Coccupancy occurred on the promoter of Leo1 in the TF1-hPRL3 as compared with TF-1 cells (Fig. 6E). JMJD2C asso-ciation to the P2 region was significantly abrogated in thePRL-3 knockdown cells, supporting the notion that JMJD2Cbinding to the Leo1 promoter region is dependent on thepresence of the PRL-3 protein (Supplementary Fig. S6E).

1.2

1

0.8

0.6

0.4

0.2

0

PRL-3

Leo1

Mock #404 #774 #780

Re

lati

ve

qu

an

tifi

ca

tio

n

Re

lati

ve

lu

cif

era

se

ac

tiv

ity

35.00

30.00

25.00

20.00

15.00

10.00

5.00

0.00Luc Emp PRL3 0.5 mg PRL3 1 mg PRL3 2 mg

TF1 TF1-PRL3

Fold increase

0.86×

0.91×

0.75×

0.96×

0.55×

0.45×

0.82×

0.83×

0.95×

H3K4me1

H3K4me2

H3K4me3

H3K79me2

H3K9me2

H3K9me3

H3K27me2

H3K27me3

H3

Rep

ressiv

eh

isto

ne m

ark

sh

isto

ne m

ark

sA

cti

ve

P1 P2 P3 P4

-439 -368 -285 -104

Leo1

H3K9me3H3K9me2

% I

np

ut

14.00

12.00

10.00

8.00

6.00

4.00

2.00

0.00

P1 P2 P3 P4

TF-1

TF-1TF1-PRL3

TF1-PRL3TF1-IgC

TF1-IgCTF1-PRL3-IgC

TF1-PRL3-IgC

10.00

8.00

6.00

4.00

2.00

0.00

% I

np

ut

P1 P2 P3 P4

**

** **

*

*

*

A B

C D

Figure 5. PRL-3 regulates Leo1 at the transcriptional level. A, Molm-14 cells were transfected with PRL-3 shRNA for 24 hours, and Leo1 and PRL-3mRNA levels were measured. Data, mean � SEM of triplicates. B, 293T cells were transfected with the respective luciferase reporter andexpression vectors, and relative luciferase activity (Firefly/Renilla) was determined. Data, mean � SEM of triplicates. C, nuclear extracts were obtainedfrom TF1 or TF1-hPRL3 cells and was immunoblotted for respective histone marks. Histone H3 was used as the loading control. Intensity of bands wasmeasured using ImageJ and provided as a fold increase using TF-1 as control. D, diagram of Leo1 promoter showing the position of four different primerpairs (P1–P4) used for qPCR. ChIP was performed in TF1 and TF1-hPRL3 cells with respective antibodies. Solid lines, ChIPs with the indicatedantibodies; dotted lines, control rabbit IgG; TF1 cells, black; TF1-hPRL3 cells, gray. Data are plotted as percentage of input. Data, mean � SEMof triplicates. �, P < 0.05; ��, P < 0.005.






Taken together, these results demonstrate that PRL-3relieves the repressive H3K9me3 mark on Leo1 promoterthrough JMJD2C, resulting in a more transcriptionally per-missive chromatin state, leading to the efficient promotionof Leo1 expression.

Leo1 suppression depletes the levels of PAF complexcomponents and affects the expression of targetgenes

The PAF complex subunits have been shown to coimmu-noprecipitate with each other. Therefore, we checkedwhether knockdown of Leo1 affects the interaction networkof the PAF complex. Interestingly, we found that loss of Leo1destabilized the PAF complex, and the mRNA levels ofPaf1, Ctr9, and Ski8 were collectively downregulated(Fig. 7A). In addition, the protein levels of Paf1 and Ctr9were also decreased with Leo1 depletion (Fig. 7B). Todetermine the functional relevance of Leo1-mediated desta-bilization of the PAF complex, we further examined theeffects on several reported PAF complex target genes. Asubset of these genes, namely Sox2, Sox4, and Tbx3, weresignificantly downregulated when Leo1 was silenced inMolm-14 cells (Fig. 7C).

To determine whether Leo1 plays an independent rolein mediating the oncogenic properties in AML or throughPAF complex as a whole, we compared the effects of Leo1,Paf1, or Ctr9 knockdown in TF1-PRL3 cells. We confirmedthe knockdown efficiency using Western Blot (Supplemen-tary Fig. S7A). Interestingly, we found that knockdown ofLeo1 impedes cell proliferation to a greater extent thanPaf1 or Ctr9 knockdown (Supplementary Fig. S7B). Inaddition, Sox2 and Sox4 expression levels were similarlyaffected in Leo1, Paf1 of Ctr9 knockdown (SupplementaryFig. S7C). These observations suggest that the entire PAFcomplex is partly involved in mediating oncogenesisthough SOX genes, but other mechanisms involving Leo1alone are present. Furthermore, when we analyzed theexpression levels of Leo1, Paf1, Cdc73, Ski8, and Ctr9 inTF1-PRL3 cells, only Leo1 showed an upregulation (Sup-plementary Fig. S7D).

Because we showed that knockdown of PRL-3 downregu-lated levels of Leo1, we further examined whether knockdownof PRL-3 similarly destabilizes the PAF complex. Interestingly,silencing PRL-3 downregulated all components of the PAFcomplex (Fig. 7D), and subsequent restoration of Leo1 levelsrescued the levels of Paf1, Ski8, and Ctr9 (Fig. 7E). These

1.80

1.60

1.40

1.20

1.00

0.80

0.60

0.40

0.20

0.00JMJD2A

JMJD2A shRNA

JM

JD

2A

sh

RN

A

JM

JD

2B

sh

RN

A

JM

JD

2C

sh

RN

A

JM

JD

2D

sh

RN

A

JMJD2B shRNA

JMJD2C shRNA

JMJD2D shRNA

JMJD2B JMJD2C JMJD2D

Rela

tive q

uan

tifi

cati

on

Mock

TF

-1

Mo

ck

Leo1

H3K9me2

H3K9me3

b-Actin

TF1-hPRL3

JMJD2C ChlP JMJD2C ChlPH3K9me3 ChlP

0.18

0.16

0.14

0.12

0.10

0.08

0.06

0.04

0.02

0

P2 P2

% I

np

ut

% I

np

ut

% I

np

ut

1.2

1

0.8

0.6

0.4

0.2

0

6.00

5.00

4.00

3.00

2.00

1.00

0.00

P1 P2 P3 P4

TF1-hPRL3TF1-hPRL3

TF-1TF1-PRL3-

TF1-hPRL3- TF1-hPRL3JMJD2C shRNA

JMJD2C shRNA

****

**

*

*

*

*

A B

C D E

Figure 6. JMJD2C regulates H3K9me3 on the Leo1 promoter. A, shRNA specificity was tested in TF-1 cells using JMJD2A, 2B, 2C, and 2D primers. Graphis plotted relative to scrambled (Mock). Data, mean � SEM of triplicates. B, TF1-hPRL3 transfected with Mock or JMJD2 shRNA and analyzed by Westernblot using the indicated antibodies. C and D, ChIP was performed using antibody against JMJD2C (C) or H3K9me3 (D) on Leo1 promoter P2. E, ChIPwas performed in TF-1 and TF1-hPRL3 cells using JMJD2C antibody on Leo1 promoter. Data shown as percentage of input. Data, mean�SEMof triplicates.�, P < 0.05; ��, P < 0.005.

Chong et al.





results further confirm that Leo1 is functionally relevant and isan important downstream mediator of PRL-3 oncogenicfunctions.

DiscussionOur present work represents the first large-scale quantita-

tive survey of proteins regulated by PRL-3 in leukemia. Weidentified several known proteins downstream of PRL-3,including stathmin, nucleolin, and translationally controlledtumor protein (6, 18, 19), which endorsed the validity of ourapproach. We also discovered additional upregulated candi-date oncogenes that were linked to carcinogenesis, such ashnRNPH1, TPI1, HSP90B1, prohibitin, as well as candidatetumor suppressors that were downregulated, such as POTE,nucleophosmin, and BTF3, which were previously not knownto be modulated by PRL-3 (Supplementary Table S1). Onthe other hand, some reported substrates of PRL-3 in solidtumor cell lines, such as CDH22, ezrin, keratin8, integrin-a1,and EF-2, were not identified in our screen, suggesting thatPRL-3 might activate both common and distinct networksin solid tumor as compared with leukemia.

From our comprehensive proteomic dataset, we gainedseveral novel insights into the functions of PRL-3. First, wefound that PRL-3 unexpectedly alters the expression levels ofa significant number of nuclear proteins (50%), although itwas reported to localize predominantly in the plasma mem-brane and cytosol due to the CAAX membrane-targetingsignal (20). Despite this, PRL-3 was also found to be able toshuttle between the nucleus and cytoplasm in myeloma andAML cells (21), by an unclear mechanism. Second, geneontology annotation by biologic process reveals a specificinvolvement of PRL-3 in pre-mRNA splicing and nucleic acidmetabolism, as well as cell-cycle regulation and cell motility.Involvement in cell motility is consistent with publishedreports in which overexpression of PRL-3 in carcinomas hasbeen linked to processes such as enhanced cell migration,invasion, and metastasis (20, 22). Notably, this is the firststudy to suggest that PRL-3 plays a role in the regulation ofRNA-related processes. Overall, our unbiased approachenables us to implicate novel candidate proteins in theoncogenic function of PRL-3, which would fundamentallyadvance our understanding of how this phosphatase con-tributes to cancer progression.

1.60

1.40

1.20

1.00

0.80

0.60

0.40

0.20

0.00

Emp #32 #33

Rela

tive q

uan

tifi

cati

on

Rela

tive q

uan

tifi

cati

on

Rela

tive q

uan

tifi

cati

on

Rela

tive q

uan

tifi

cati

on

PAF1

Leo1

CDC73

CTR9

SKI8

* *

** *

**

Leo1

shRNA

Mock

#32#33

Ctr9

Paf1

Leo1

b-Actin

1.20

1.00

0.80

0.60

0.40

0.20

0.00

1.20

1.00

0.80

0.60

0.40

0.20

0.00

Sox2 Sox4 Tbx3

****

*

Emp

Leo1 shRNA

Emp #774 #780

** ** ** **

**

*

*

PAF1

PAF1LEO1

LEO1CDC73

CDC73

CTR9

CTR9

SKI8

SKI8

PRL-3**

**3.00

2.50

2.00

1.50

1.00

0.50

0.00Emp #774 Rescue #780 Rescue

A B C

D E

Figure 7. Leo1 knockdown destabilizes the PAF complex. A and B, Molm-14 cells were transfected with either Leo1 shRNA or control shRNA (Emp), andqRT-PCR (A) or Western blot (B) was done for the expression of the PAF complex components. Data, mean � SEM of triplicates. C, qRT-PCRanalyses of Sox2, Sox4, and Tbx3 after Leo1 knockdown in Molm-14 cells. Data, mean � SEM of triplicates. D, Molm-14 was transfected with PRL-3shRNA and analyzed for PAF components. Data, mean � SEM of triplicates. E, Molm-14 was cotransfected with PRL-3 shRNA and Leo1 expressionconstruct and analyzed for PAF components. Data, mean � SEM of triplicates. �, P < 0.05; ��, P < 0.005.






In the recent years, several pivotal articles have proposedfunctions of the PAF complex in development and cancer, andthe PAF complex was found to be substrates of anotheroncogenic phosphatase, SHP2 (23–26). In addition, the PAFcomplex is essential for leukemogenesis by mixed-lineageleukemia fusion proteins through inducing the expression ofHox genes, thus maintaining a haematopoietic stem cell(HSC)–like signature and resistance to differentiation (15).These led us to focus on the involvement of Leo1 and the PAFcomplex as the downstream mediators of PRL-3 activities.Although characterization of the individual subunits of thePAF complex revealed opposing roles in cancer (24), ourcurrent work suggests that Leo1 is a proto-oncogene in leu-kemia, and the levels of Ski8, Ctr9, and Paf1 subunits dependedon Leo1 levels (Fig. 7A and B), supporting the notion of aninterdependency within the PAF complex.

One important hallmark of AML is the acquisition of self-renewal capacity in the leukemic blast cells (27) throughthe disruption of self-renewal pathways such as Hox andWnt/b-catenin. Interestingly, the PAF complex has beenfound to augment the expression of HOXA9 (15), and inanother independent report, Leo1 has been found to asso-ciate directly with b-catenin (28). In addition, the PAFcomplex has also been reported to regulate directly keypluripotency genes in embryonic stem cells. Indeed, whenwe knocked down Leo1, several pluripotency genes, such asSox2, Sox4, and Tbx3, were downregulated (Fig. 7C). Thederegulation of Sox genes, especially Sox4, has been shown tobe an important collaborating driver of myeloid leukemo-genesis in several transgenic murine models of leukemia(29). Thus, our study suggests that PRL-3 could exert onco-genic properties in AML through Leo1 and PAF complex–mediated deregulation of critical pluripotency genes previ-ously implicated in leukemogenesis.

There were few reports on how PRL-3 regulates its targets,in part due to the challenges in identifying the genuinetargets of PRL-3. In our study, we revealed a novel mech-anism of the transcriptional involvement of PRL-3 in regu-lating Leo1. Trimethylation on H3K9 is a strong repressiveepigenetic mark generally associated with heterochromatinmaintenance and transcriptional repression, and deregula-tion of H3K9me3 levels has been shown to cause aberrantgene expression, contributing to the development of cancer(30, 31). We found global demethylation of H3K9me3 in TF1-hPRL3 cells, suggesting that alterations in a histone-modi-fying enzyme lead to these widespread epigenetic changes(32). JMJD2C, originally identified as GASC1, is a gene locatedon chromosome 9p, and is frequently amplified in cancers

(30). Our results identify JMJD2C as the main enzymeinvolved in the demethylation of H3K9me3 on the Leo1promoter in TF1-hPRL3 cells, thereby enhancing the expres-sion of Leo1. Interestingly, we noticed an increase in JMJD2Coccupancy on the promoter region of Leo1, and this wasdependent on PRL-3 expression levels (Fig. 6E). How PRL-3affects the binding of JMJD2C or possibly other histonedemethylase on the promoters of target genes remains tobe determined. Importantly, our results demonstrated apossible interface between the activities of a phosphataseand a histone demethylase in gene regulation.

In conclusion, our study unravels downstream signalingpathways of PRL-3, identifying Leo1 as an important down-stream target. Taken together, this study provides novelinsights into the mechanisms by which PRL-3 promotes themalignant phenotype in AML.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: P.S.Y. Chong, J. Zhou, W.J. ChngDevelopment of methodology: P.S.Y. Chong, J. Zhou, S.K. Sze, W.J. ChngAcquisition of data (provided animals, acquired and managedpatients, provided facilities, etc.): J. Zhou, L.-L. Cheong, J. Qian, S.K. Sze,W.J. ChngAnalysis and interpretation of data (e.g., statistical analysis, biosta-tistics, computational analysis): P.S.Y. Chong, J. Zhou, J. Qian, T. Guo, S.K.Sze, W.J. ChngWriting, review, and or revision of the manuscript: P.S.Y. Chong, J. Zhou,W.J. ChngAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): L.-L. Cheong, S.-C. Liu, J. Qian, Q. Zeng,W.J. ChngStudy supervision: J. Zhou

AcknowledgmentsThe authors thank professors Daniel G. Tenen, H. Phillip Koeffler, and

assistant professor Motomi Osato for helpful suggestions.

Grant SupportW.J. Chng is supported by a National Medical Research Council (NMRC)

Clinician Scientist Investigator award. This work is partly supported by aSingapore Cancer Syndicate Grant. This research is also supported by theNational Research Foundation Singapore and the Singapore Ministry ofEducation under the Research Centers of Excellence initiative. J. Zhou issupported by an NMRC Clinician-Scientist IRG grant CNIG11nov38.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicate thisfact.

Received August 13, 2013; revised February 7, 2014; accepted March 6, 2014;published OnlineFirst March 31, 2014.

References1. Zeng Q, Si X, Horstmann H, Xu Y, Hong W, Pallen CJ. Prenylation-

dependent association of protein-tyrosine phosphatases PRL-1, -2,and -3 with the plasma membrane and the early endosome. J BiolChem 2000;275:21444–52.

2. Al-Aidaroos AQ, Zeng Q. PRL-3 phosphatase and cancer metastasis.J Cell Biochem 2010;111:1087–98.

3. Kato H, Semba S, Miskad UA, Seo Y, Kasuga M, Yokozaki H. Highexpression of PRL-3 promotes cancer cell motility and liver metastasis

in human colorectal cancer: a predictive molecular marker ofmetachronous liver and lung metastases. Clin Cancer Res 2004;10:7318–28.

4. Polato F, Codegoni A, Fruscio R, Perego P, Mangioni C, Saha S, et al.PRL-3 phosphatase is implicated in ovarian cancer growth. ClinCancer Res 2005;11:6835–9.

5. Radke I, Gotte M, Kersting C, Mattsson B, Kiesel L, Wulfing P.Expression and prognostic impact of the protein tyrosine


Chong et al.




phosphatases PRL-1, PRL-2, and PRL-3 in breast cancer. Br J Cancer2006;95:347–54.

6. Chu ZH, Liu L, Zheng CX, Lai W, Li SF, Wu H, et al. Proteomic analysisidentifies translationally controlled tumor protein as a mediator ofphosphatase of regenerating liver-3-promoted proliferation, migrationand invasion in human colon cancer cells. Chin Med J (Engl) 2011;124:3778–85.

7. Jiang Y, Liu XQ, Rajput A, Geng L, Ongchin M, Zeng Q, et al. Phos-phatase PRL-3 is a direct regulatory target of TGFbeta in colon cancermetastasis. Cancer Res 2011;71:234–44.

8. Saha S, Bardelli A, Buckhaults P, Velculescu VE, Rago C, Croix BS,et al. A phosphatase associated with metastasis of colorectal cancer.Science 2001;294:1343–6.

9. Bardelli A, SahaS,Sager JA,RomansKE,XinB,MarkowitzSD, et al. PRL-3 expression in metastatic cancers. Clin Cancer Res 2003;9:5607–15.

10. Wang H, Vardy LA, Tan CP, Loo JM, Guo K, Li J, et al. PCBP1suppresses the translation of metastasis-associated PRL-3 phospha-tase. Cancer Cell 2010;18:52–62.

11. Bessette DC, Qiu D, Pallen CJ. PRL PTPs: mediators and markers ofcancer progression. Cancer Metastasis Rev 2008;27:231–52.

12. Guo K, Li J, Tang JP, Tan CBP, Hong CW, Al-Aidaroos AQ, et al.Targeting intracellular oncoproteins with antibody therapy or vacci-nation. Sci Transl Med 2011;3:99ra85.

13. Zhou J, Bi C, Chng WJ, Cheong LL, Liu SC, Mahara S, et al. PRL-3, ametastasis associated tyrosine phosphatase, is involved in FLT3-ITDsignaling and implicated in anti-AML therapy. PLoS ONE 2011;6:e19798.

14. Beekman R, Valkhof M, Erkeland SJ, Taskesen E, Rockova V, PeetersJK, et al. Retroviral integration mutagenesis in mice and comparativeanalysis in human AML identify reduced PTP4A3 expression as aprognostic indicator. PLoS ONE 2011;6:e26537.

15. MunteanAG, Tan J, Sitwala K,HuangY,Bronstein J,Connelly JA, et al.The PAF complex synergizes with MLL fusion proteins at HOX loci topromote leukemogenesis. Cancer Cell 2010;17:609–21.

16. LiuY, ZhengP, Liu Y, Ji T, Liu X, YaoS, et al. An epigenetic role for PRL-3 as a regulator of H3K9 methylation in colorectal cancer. Gut2013;62:571–81.

17. Shin S, Janknecht R. Diversity within the JMJD2 histone demethylasefamily. Biochem Biophys Res Commun 2007;353:973–7.

18. Semba S, Mizuuchi E, Yokozaki H. Requirement of phosphataseof regenerating liver-3 for the nucleolar localization of nucleolinduring the progression of colorectal carcinoma. Cancer Sci 2010;101:2254–61.

19. ZhengP, LiuYX,Chen L, Liu XH,XiaoZQ, ZhaoL, et al. Stathmin, a newtarget of PRL-3 identified by proteomic methods, plays a key role inprogression and metastasis of colorectal cancer. J Proteome Res2010;9:4897–905.

20. Zeng Q, Dong JM, Guo K, Li J, Tan HX, Koh V, et al. PRL-3 and PRL-1promote cell migration, invasion, and metastasis. Cancer Res2003;63:2716–22.

21. Fagerli UM, Holt RU, Holien T, Vaatsveen TK, Zhan F, Egeberg KW,et al. Overexpression and involvement in migration by the metastasis-associated phosphatase PRL-3 in human myeloma cells. Blood2008;111:806–15.

22. Fiordalisi JJ, Keller PJ, Cox AD. PRL tyrosine phosphatases regulaterho family GTPases to promote invasion and motility. Cancer Res2006;66:3153–61.

23. Squazzo SL, Costa PJ, Lindstrom DL, Kumer KE, Simic R, JenningsJL, et al. The Paf1 complex physically and functionally associateswith transcription elongation factors in vivo. EMBO J 2002;21:1764–74.

24. Chaudhary K, Deb S, Moniaux N, Ponnusamy MP, Batra SK. HumanRNA polymerase II-associated factor complex: dysregulation in can-cer. Oncogene 2007;26:7499–507.

25. Wang P, Bowl MR, Bender S, Peng J, Farber L, Chen J, et al. Parafi-bromin, a component of the human PAF complex, regulates growthfactors and is required for embryonic development and survival in adultmice. Mol Cell Biol 2008;28:2930–40.

26. Takahashi A, Tsutsumi R, Kikuchi I, Obuse C, Saito Y, Seidi A, et al.SHP2 tyrosine phosphatase converts parafibromin/Cdc73 from atumor suppressor to an oncogenic driver. Mol Cell 2011;43:45–56.

27. Tenen DG. Disruption of differentiation in human cancer: AML showsthe way. Nat Rev Cancer 2003;3:89–101.

28. Mosimann C, Hausmann G, Basler K. Parafibromin/Hyrax activatesWnt/Wg target gene transcription by direct association with beta-catenin/Armadillo. Cell 2006;125:327–41.

29. Sandoval S, Kraus C, Cho EC, Cho M, Bies J, Manara E, et al. Sox4cooperates with CREB in myeloid transformation. Blood 2012;120:155–65.

30. Cloos PA, Christensen J, Agger K, Maiolica A, Rappsilber J, Antal T,et al. The putative oncogene GASC1 demethylates tri- and dimethy-lated lysine 9 on histone H3. Nature 2006;442:307–11.

31. Bannister AJ, Kouzarides T. Regulation of chromatin by histone mod-ifications. Cell Res 2011;21:381–95.

32. Fullgrabe J, Kavanagh E, Joseph B. Histone onco-modifications.Oncogene 2011;30:3391–403.






2014;74:3043-3053. Published OnlineFirst March 31, 2014.Cancer Res Phyllis S.Y. Chong, Jianbiao Zhou, Lip-Lee Cheong, et al. in Acute Myelogenous LeukemiaLEO1 Is Regulated by PRL-3 and Mediates Its Oncogenic Properties

Updated version

10.1158/0008-5472.CAN-13-2321doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2014/04/01/0008-5472.CAN-13-2321.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/74/11/3043.full#ref-list-1

This article cites 32 articles, 14 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/74/11/3043.full#related-urls

This article has been cited by 3 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/74/11/3043To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-13-2321

http://cancerres.aacrjournals.org/content/suppl/2014/04/01/0008-5472.CAN-13-2321.DC1

http://cancerres.aacrjournals.org/content/74/11/3043.full#ref-list-1

http://cancerres.aacrjournals.org/content/74/11/3043.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/74/11/3043


LEO1 Is Regulated by PRL-3 and Mediates Its Oncogenic ... · Molecular and Cellular Pathobiology...

Documents

Transcript of LEO1 Is Regulated by PRL-3 and Mediates Its Oncogenic ... · Molecular and Cellular Pathobiology...