Human Cytomegalovirus US28 Found in Glioblastoma Promotes ... › content › canres › ... ·...

Molecular and Cellular Pathobiology

Human Cytomegalovirus US28 Found in GlioblastomaPromotes an Invasive and Angiogenic Phenotype

Liliana Soroceanu1, Lisa Matlaf1, Vladimir Bezrookove1, Loui Harkins3, Roxanne Martinez1, Mary Greene1,Patricia Soteropoulos4, and Charles S. Cobbs1,2

AbstractHuman cytomegalovirus (HCMV) infections are seen often in glioblastoma multiforme (GBM) tumors, but

whether the virus contributes to GBMpathogenesis is unclear. In this study, we explored an oncogenic role for theG-protein–coupled receptor-like protein US28 encoded by HCMV that we found to be expressed widely in humanGBMs. Immunohistochemical and reverse transcriptase PCR approaches established that US28 was expressed inapproximately 60% of human GBM tissues and primary cultures examined. In either uninfected GBM cells orneural progenitor cells, thought to be the GBM precursor cells, HCMV infection or US28 overexpression wassufficient to promote secretion of biologically active VEGF and to activate multiple cellular kinases that promoteglioma growth and invasion, including phosphorylated STAT3 (p-STAT3) and endothelial nitric oxide synthase(e-NOS). Consistent with these findings, US28 overexpression increased primary GBM cell invasion in Matrigel.Notably, this invasive phenotype was further enhanced by exposure to CCL5/RANTES, a US28 ligand, associatedwith poor patient outcome in GBM. Conversely, RNA interference–mediated knockdown of US28 in humanglioma cells persistently infected with HCMV led to an inhibition in VEGF expression and glioma cell invasion inresponse to CCL5 stimulation. Analysis of clinical GBM specimens further revealed that US28 colocalized in situwith several markers of angiogenesis and inflammation, including VEGF, p-STAT3, COX2, and e-NOS. Takentogether, our results indicate that US28 expression fromHCMV contributes to GBM pathogenesis by inducing aninvasive, angiogenic phenotype. In addition, these findings argue that US28–CCL5 paracrine signaling maycontribute to glioma progression and suggest that targeting US28 may provide therapeutic benefits in GBMtreatment. Cancer Res; 71(21); 6643–53. �2011 AACR.

Introduction

Human cytomegalovirus (HCMV) is a beta-herpesvirus thatcan cause life-threatening infections in human fetuses andimmunocompromised individuals. It is a major cause of con-genital brain infection and disability in humans. Our laboratoryfirst reported the association of HCMV infection with humanglioblastomas (GBM; ref. 1). These findings have been con-firmed by other groups (2), and clinical trials are currentlyunderway for the treatment of GBM with anti-HCMV agentsand immunotherapy approaches. A growing body of evidence

suggests that HCMV infection of malignant glioma does notsimply represent an epiphenomenon but rather expression ofHCMV gene products in tumor cells and the tumor microen-vironment directly impacts tumor progression. Several HCMVgene products have been found to have mutagenic and trans-forming potential (3). We previously showed that the HCMVIE1 gene product promotes GBM mitogenicity by interferingwith Rb, p53, and AKT signaling pathways (4) and that HCMVenvelope glycoprotein B can directly activate oncogenic sig-naling pathways through activation of the PDGFR-a receptortyrosine kinase (5).

US28 is anHCMV-encodedG-protein–coupled receptor thatis a homologue of the human CCR1 chemokine receptor. US28is constitutively active andmay be further activated by bindingof several ligands: SDF-1, CCL2/MCP-1, CCL5/RANTES, andCX3CL1/Fraktalkine (6). US28 has properties of a viral onco-gene, because ectopic expression of US28 can induce a proan-giogenic and transformed phenotype in vivo via activation ofthe NF-kB and COX2 signaling pathways (7). A recent reportshowed that US28 induces interleukin-6 (IL-6) and VEGFthrough NF-kB activation, resulting in potent activation ofthe STAT-3 transcriptional activator in NIH 3T3 mouse fibro-blasts (8).

We sought to determine whether US28 could (i) influenceearly events that lead to gliomagenesis in neural precursorcells and (ii) promote key oncogenic features in established

Authors' Affiliations: 1California PacificMedical Center Research Institute;2Department of Neurological Surgery, University of California, San Fran-cisco, San Francisco, California; 3Birmingham Veterans AdministrationHospital, Pathology Section, Birmingham, Alabama; and 4Institute of Geno-mic Medicine, UMDNJ-New Jersey Medical School, Newark, New Jersey

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

Corresponding Authors: Liliana Soroceanu, California Pacific MedicalCenter Research Institute, 475 Brannan Street, Suite 220, SanFrancisco, CA 94107. Phone: 415-600-1770; Fax: 415-600-1725; E-mail:[email protected]; and Charles S. Cobbs, California Pacific MedicalCenter Research Institute, 475 Brannan Street, Suite 220, San Francisco,CA 94107. E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-11-0744

�2011 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 6643

on August 1, 2020. © 2011 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 7, 2011; DOI: 10.1158/0008-5472.CAN-11-0744

http://cancerres.aacrjournals.org/

glioblastoma cells. We hypothesized that US28 expression inneural precursor cells (NPC), possibly the glioma "cell oforigin", may promote the pathogenesis of GBM. We evalu-ated changes that occur in gene expression patterns and inthe angiogenic and invasion pathways in adult NPCs, wheninfected by HCMV or overexpressing US28. Next, we assessedthe changes in the invasive and angiogenic phenotypes ofprimary GBM cultures induced by US28 overexpression.Finally, we conducted loss-of-function studies in humanGBMs persistently infected with HCMV to show the spec-ificity of US28-induced pathogenesis.

Materials and Methods

Cell cultureU251 and U87 cell lines were obtained from American Type

Culture Collection (ATCC) and grown in DMEM/Ham's F-12þ10% FBS. Primary glioblastoma/neural precursor cell–derivedcultures were generated with tissue from surgical resections atthe California PacificMedical Center obtained according to theInstitutional Review Board–approved protocol. Tissues weredissociated by enzymatic and mechanical dissociation aspreviously described. Single-cell suspensions were culturedwith neural basal medium þ N2 supplement, 20 ng/mL epi-dermal growth factor (EGF), 20 ng/mL basic fibroblast growthfactor, and 1 mg/mL laminin as previously described (9). ForELISA for VEGF experiments and tube formation assays, cellswere cultured in the absence of FBS or growth factors at least48 hours prior to media collection. The NPC cell line wasderived from the hippocampus tissue removed from a patientwith intractable epilepsy. Cells were characterized by immu-nofluorescence and found positive for Nestin, GFAP, Tuj1, andOlig2. All experiments were carried out on passages 2 to 5 fromthe NPC culture. Human umbilical vein endothelial cells(HUVEC) were obtained from Invitrogen and grown in thecomplete endothelial cell growth media recommended by themanufacturer.

US28 expression vectorsThe Ad-US28 and Ad-Control adenoviruses were a gift from

Dr. Dan Streblow, Oregon Health & Science University. ThepcDEF-US28 plasmid was a gift from Dr. Martine Smit. TheUS28 insert was excised from the pcDEF plasmid and clonedinto the pLXSN vector (4). Retroviruses were produced andused to infect glioma cells as previously described by ourlaboratory (4).

VirusesThe Towne and AD169 HCMV strains were obtained from

ATCC and grown in human embryonic fibroblasts, as previ-ously described (10). The TR virus strain was a gift fromDr. LeeFortunato, University of Idaho (Moscow, ID).

Knockdown experiments using siRNA to US28US28 knockdown was achieved with 2 siRNA oligonucleotide

duplexes custom synthesized by Dharmacon. The sensesequences for the 2 siRNAs are as follows: CGACGGAGUUUGA-CUACGAUU(1) and CUCACAAAUUACCGUAUU(2). Experi-

ments were carried out with each siRNA individually and the2 duplexes were combined. As a negative control, nontargetingcontrol pool fromDharmacon (D-001810-10-05)was used. Effec-tive protein knockdown was verified at 48 and 72 hours post-transfection and prior to functional assays, as described later.

Fluorescencemeasurements to quantifyUS28 expressionlevels

Images were taken at fixed exposure times with an AxioImage Z2 microscope (Zeiss). The fluorescence intensities,from at least 100 cells, were quantified with ImageJ software;plots representing cumulative distribution ofmean pixel inten-sity for various conditions are shown. The Kolmogorov–Smir-nov test was used to determine whether the measured differ-ences were statistically significant.

Expression profiling using the HCMV DNA array and theAffymetrix Gene ST1 array

Total RNA was isolated and the quality verified as describedlater for reverse transcriptase PCR (RT-PCR). The RNA wasprocessed for microarray hybridization at the Center forApplied Genomics, UMDNJ-New Jersey Medical School. TheHCMV arrays were printed and processed as described previ-ously (11). Briefly, the array contains 65-mer oligonucleotidesrepresenting 194 predicted open reading frames (ORF) of theHCMV strain AD169, 19 oligonucleotides for ORFs in theToledo strain that are not found in AD169, and 44 humangenes as controls. Total RNA (3 mg) was reversed transcribed tocDNA using SuperScript II RT in the presence of cyanine-3(Cy3) or cyanine-5 (Cy5) dUTP. The labeled cDNA was purifiedand hybridized to the arrays at 58�C for 16 hours. The slideswere scanned with an Axon 4200AL scanner, and the imageswere processed with GenePix Pro 6.1. A normalization factorwas calculated using 36 human control genes (11) by dividingthe median intensity of the Cy5 signal by the median intensityof Cy3 signal of the controls. The data were normalized bymultiplying the Cy3 signal of each spot by the normalizationfactor. The ratio of the Cy5median intensity to the Cy3medianintensity was determined for each spot and the average ratiodetermined for the replicate spots. The accession number fordata from both Affymetrix and HCMV platforms is GSE31142.

HUVEC tube formation assayGeltrex (Invitrogen #12760-013) was obtained from Invitro-

gen and thawed overnight at 4�C. One hundred microliters ofGeltrex per well was placed on the bottom of 24-well culturedishes and allowed to solidify at 37�C for 30 minutes. HUVECswere detached with EDTA and resuspended in endothelial cellmedium supplemented with various growth factors or condi-tioned media at 40,000 cells/200mL per well. Tubes wereallowed to form for 8 to 10 hours, and cells were visualizedwith aNikon Inverted Eclipse TE-2000Emicroscope,fittedwitha CCD Cascade II camera. NIS Elements AR3.0 was used toacquire images, which were further processed in Photoshop.

Statistical data analysisSignificant differences were determined by ANOVA or the

unpaired Student t test, where suitable. Bonferroni–Dunn post

Soroceanu et al.

Cancer Res; 71(21) November 1, 2011 Cancer Research6644




hoc analyses were conducted when appropriate. The values ofP < 0.05 defined statistical significance.Additional methods are available as Supplementary

Information.

Results

US28 is endogenously expressed in human GBMsUS28 protein expression in human glioblastomas was

assessed by immunofluorescence analysis of primary glioblas-toma-derived cultures and immunohistochemical analysis ofparaffin-embeddedtissues fromseveralGBMspecimens, includ-ingsomethatwereused togenerate theprimarycultures. reversetranscriptase PCR for US28, HCMV UL56 (a DNA packagingessential viral gene), andRab14 (humanhousekeepinggene)wasdoneusingRNA isolated fromsnap-frozen tissues fromthe samecases. Figure 1A shows an example of immunofluorescence

analysis of primary GBM cells that exhibit cytoplasmic andmembrane staining for the US28 antigen. Preincubation of theprimary antibody with excess US28 blocking peptide showedspecificity of immunostaining (Fig. 1B).

As shown in Fig. 1C, US28 expression was detected inparaffin-embedded GBM biopsy specimens. Figure 1D showsspecificity of staining, using the US28 blocking peptide inexcess, as described earlier. Sections from the same sampleshow abundant staining for VEGF (Fig. 1E) and COX2 (Fig.1F), suggesting the presence of enhanced angiogenesis andinflammation in and around the US28-positive tumor cells.Colocalization of US28 and VEGF in another case of primaryglioblastoma is shown in Supplementary Fig. S1. The spec-ificity of the US28 antibody was established by comparingimmunostaining of cells that were mock-infected, HCMV-infected, or ectopically expressing US28 (Supplementary Fig.S2). To confirm that HCMV US28 mRNA was likewise

Figure 1. HCMV US28 transcriptand protein are expressed in humanGBMs. A and B, primary GBM-derived cultureswere processed forUS28 immunofluorescence in theabsence (A) or presence (B) of ablocking peptide. Nuclei arecounterstained with propidiumiodide. Bar, 100 mm. C–F,consecutive (5 mm) paraffinsections obtained from a differentGBM patient sample wereprocessed for US28 (C and D),VEGF (E), and COX2 (F)immunohistochemistry.Counterstaining, hematoxylin. Bar,100 mm. G, reverse transcriptasePCR for US28 was done usingcDNA from several GBM cases.HCMV-infected neural precursorcells (NPC þ CMV) served aspositive controls. Several casesshow a US28 band of the correctsize. HCMV UL56 detection is alsoshown. Rab14 was used to verifyequal loading.NC, negative control.

A B

C D

E

G

500 bp

N.C

.

NP

Cs C

MV

CP

MC

1

CP

MC

2

CP

MC

3

CP

MC

6

CP

MC

9

CP

MC

12

CP

MC

14

CP

MC

19

CP

MC

20

CP

MC

23

CP

MC

25

CP

MC

26

UL5

6U

S28

Rab14

NP

Cs m

ock

250 bp

500 bp

250 bp

500 bp

250 bp

F

HCMV US28 Promotes GBM Invasion and Angiogenesis

www.aacrjournals.org Cancer Res; 71(21) November 1, 2011 6645




expressed in human GBM specimens, we carried out reversetranscriptase PCR on RNA extracted from GBM biopsyspecimens from several different patients. Uninfected NPCsshowed no evidence of the amplified US28 gene product, oranother conserved HCMV gene product UL56 (Fig. 1G). Incontrast, we detected amplified US28 RNA transcripts in theprimary GBM biopsy specimens from several patients,including a case found positive by immunohistochemistry(shown in Fig. 1C–F). All amplified US28 reverse transcrip-tase PCR products were sequenced to confirm specificity toHCMV, and unique gene polymorphisms were identified inseveral specimens, indicating that no cross-contaminationof laboratory or PCR specimen occurred (C-terminalsequences alignment is provided in the SupplementaryInformation). Additional GBM and control brain tissueswere immunostained for US28, COX2, VEGF, phospho-STAT3 (p-STAT3), and endothelial nitric oxide (e-NOS; Sup-plementary Table S1). Of the 35 different brain tissuesscreened, 53% were positive for US28 by reverse transcrip-tase PCR and 65% were positive by immunohistochemistry;there was more than 90% concordance in the results show-ing US28 detection when both approaches were used (Sup-plementary Table S1).

HCMV infection of NPCs induces expression of US28 andCCL5, which together promote glioma invasiveness

To understand the role HCMV US28 might play in glioma-genesis, we first wished to ascertain that US28 is expressedduring HMCV infection of human adult NPCs, the purportedcells of origin of adult GBM. NPCs were infected with HCMV[Towne and TR strains; multiplicity of infection (MOI) ¼ 1] ormock infected. Total RNA was harvested at 72 hours, andHCMV gene expression was assayed with a custom-madeoligonucleotide microarray representing all the predictedORFs for HCMV (11). The same samples were profiled withhuman Affymetrix DNA arrays. As shown in Fig. 2A, US28 wasamong the most abundantly expressed HCMV transcriptsfollowing infection with either viral strain. Interestingly, oneof most upregulated human transcripts was the chemokineCCL5/RANTES (Fig. 2B, arrow). Although US28 can act as aconstitutively active receptor, CCL5 is a bona fide ligand forUS28 and can further stimulate US28 signaling, suggesting thatUS28 and CCL5/RANTES coexpression might induce a potentautocrine signaling loop. To determine whether expression ofCCL5 is a relevant biomarker for GBM, we analyzed theREMBRANDT GBM database. We determined that CCL5expression levels were inversely correlated with survival inhuman glioblastomas (Fig. 2C). Analysis of previously charac-terized glioblastoma molecular subclasses (12) showed thatCCL5 expression levels are elevated in the "mesenchymal"GBMs, characterized by poor patient outcome (ref. 12; Sup-plementary Fig. S3).

To assess the effects of US28 expression on glioma inva-siveness, we carried out Matrigel invasion assays comparingLXSN with US28-LXSN–transduced U251 and U87 glioma cellsand 2 primary glioma cultures, which had no detectable HCMVtranscripts. US28 overexpression resulted in an approximately30% increase in the invasiveness of all glioma cell lines tested

(Fig. 2D). The presence of 50 ng/mL recombinant human CCL5in the bottom chamber further enhanced invasiveness ofglioma cells and primary GBM cultures by 50% to 60%, asshown in Fig. 2D. These data show that CCL5, which isupregulated by HCMV infection, can augment US28-inducedglioma cell invasion.

To establish the specificity of US28 effects on glioma cellinvasion, we used a siRNA approach to knockdown US28expression in a well-characterized human glioma cell line,U87 (13), persistently infected with HCMV (SupplementaryFig. S4). US28 protein levels were measured by fluorescenceintensity measurements of cells processed for US28 immuno-fluorescence. US28 siRNA1 induced an approximately 40%US28 knockdown, whereas siRNA2 induced approximately60% US28 knockdown (Supplementary Fig. S5). When usedtogether, siRNA1þ2 induced approximately 80% US28 knock-down (Supplementary Fig. S5). We used a CCL5-neutralizingantibody to distinguish between US28 constitutive activity andthe response to the CCL5 ligand secreted by human gliomacells. Figure 2E shows that CCL5 levels were significantly(�75%) inhibited in U87 cells by preincubation with aCCL5-neutralizing antibody (20 ng/mL, 12 hours), regardlessof the presence of HCMV or US28. Although US28 KD had noeffect in uninfected U87 cells, Matrigel invasion of HCMV-positive U87 cells was inhibited by approximately 20% by US28siRNA1 or 2 used alone and by 30% when the 2 siRNAs wereused together (Fig. 2F and Supplementary Fig. S5). Pretreat-ment with CCL5-neutralizing antibody inhibited glioma cellinvasion by approximately 30% to 35% and the use of bothUS28knockdown and CCL5 neutralization did not further increasethis effect (Fig. 2F and Supplementary Fig. S5). US28 knock-down a primary GBM culture, confirmed to be HCMV positive,resulted in inhibition of tumor cell invasion by approximately35%, both baseline and in response to CCL5 stimulation(Supplementary Fig. S6).

US28 activates multiple oncogenic pathwaysin human NPCs

To determine additional oncogenic pathways activated byHCMV infection/US28 expression in NPCs, we used a phos-phor-kinase human array (R&D Systems) embedded withantibodies specific for multiple phosphoproteins (Fig. 3A andB). Pathways associated with glioma progression and invasion,including p-STAT3, AKT, ERK1/2, FAK, Src, and eNOS, weresignificantly activated by both whole virus infection and US28overexpression in NPCs (Fig. 3C). Immunofluorescence anal-yses of US28 overexpressing NPCs confirmed upregulation ofCOX2, VEGF, p-STAT3, and e-NOS (Fig. 3D). e-NOS levels,which are elevated in gliomas, correlate with increased tumoraggressiveness (14, 15). In addition to its proangiogenic role,e-NOS mediates production of nitric oxide, which was shownto induce the growth of glioma-initiating cells (16). This is thefirst report documenting that HCMV US28 induces e-NOSactivation, which contributes to glioma pathogenesis.

Using Western blotting and immunofluorescence, we con-firmed that US28 induces p-STAT3 in neural precursor cells(Supplementary Fig. S7). STAT3 activation is critical for NPCmalignant transformation toward a mesenchymal GBM

Soroceanu et al.





phenotype (17), suggesting that US28-induced activation ofp-STAT3 may contribute to gliomagenesis. Consistent with arecent report, we also found that US28 and p-STAT3 colocalizein primary glioblastomas in situ (Supplementary Fig. S8),which would explain why HCMV-positive glioma cells exhibitactivation of the STAT3 pathway, implicated in promotingimmunosuppression, maintenance of glioma stem cells, andtumor progression (8, 18).

US28 promotes GBM angiogenesisWe next investigated whether US28 can modulate VEGF

levels in neural precursor and glioma cells by imunofluor-

escence and ELISA. Figure 4A shows that VEGF is signifi-cantly upregulated in US28-expressing NPCs. VEGF levelswere measured in 4 different cell types (NPC, U251 and U87glioma cell lines, and a primary GBM–derived line) by ahighly sensitive ELISA. Seventy-two hours following infec-tion with either Towne or TR HCMV strain, or US28 over-expression, VEGF was induced more than 2-fold in allcell types tested (Fig. 4B). US28 overexpression alone wassufficient to induce equivalent levels of VEGF expression tothose found after whole HCMV infection, suggestingthat US28 may play a predominant role in the HCMV-induced VEGF secretion. Remarkably, NPCs, which are

Figure 2. US28–CCL5 signalingpromotes glioblastomainvasiveness. A,NPCs infectedwithTowne and TR HCMV strains(MOI ¼ 1, 72 hours) were profiledwith an HCMV DNA microarraycontaining all predicted ORFs forAd169/Toledo strains. Expressionlevels of HCMV transcripts aredisplayed as fold increase overuninfected control. B, RNA fromHCMV-treated and control NPCswere profiled with Affymetrix Gene1.0 ST DNA arrays. The heatmapshows the 30most upregulated and30 most downregulated humantranscripts in HCMV-infectedNPCsversus mock. CCL5 was inducedmore than 40-fold by HCMVtreatment (arrow). C, Kaplan–Meiercurves showing the relationshipbetween levels of CCL5 transcriptand survival probability in patientswith glioblastoma (log-rank P valueupregulated vs. all other samples,P ¼ 0.001523, REMBRANDTdatabase, National CancerInstitute). D, human glioma cells(U251 and U87) and 2 primaryglioblastoma-derived cultures(designated GBM#1 and GBM#2)transfected with US28 or controlvector were subjected to Matrigelinvasion assays in the absence orpresence of CCL5 (50 ng/mL).��, P < 0.005, ANOVA. E, CCL5levels measured by ELISA in mock-treated U87 cells or HCMV-infectedwith or without CCL5-neutralizingantibody. ��, P < 0.005, ANOVA. F,mock-treated and HCMV-infectedU87 cells were subjected toMatrigel invasion assays. Meannumber of cells per filter is shownfor each condition. ��, P < 0.005,ANOVA. US28 knockdown wasachieved with 2 siRNA duplexes incombination (siRNA1þ2). Datafrom 1 representative experimentare shown. Each experiment wascarried out in triplicate, andexperimentswere repeated 3 times.

A

B C

E F

D

US28 Towne

Towne

TR

TR

CCL5Color key

1.00

0.75

0.50

0.25

0.00

100

80

60

40

20

0

80

70

60

50

40

30

20

10

0

25

20

15

10

5

0

Value0 20 40 60

pp65

gB

lE1

Pro

ba

bilit

y o

f s

urv

iva

l%

In

cre

ase f

rom

co

ntr

ol

Avera

ge c

ell

nu

mb

er

CC

L5/r

an

tes (

ng

/mL

)

Days of study

1,000

Baseline

U251 U87

U87 U87 + Towne U87 + TRU87 U87 + Towne U87 + TR

GBM#1 GBM#2

****

**

**

*

** ** **

*

* * * * * *

**

**

**

**

CCL5

CCL5 MabCCL5 Mab + US28 siRNA1+2

US28 siRNA1+2Control

CCL5 MabCCL5 Mab + US28 siRNA1+2

US28 siRNA1+2Control

CCL5 downregulated

CCL5 upregulated

All gliomas

3,000 6,000 7,500

TolU

L148

HCM

VUS28

HCM

VUS12

HCM

VUL1

32

TRL1

4C

HCM

VUL1

18

UL4

2A

HCM

VIRS1

HCM

VUL5

HCM

VUL2

1

HCM

VUS22

HCM

VUS8

HCM

VUL5

8

HCM

VUL4

9

HCM

VUL5

5

HCM

VUL1

23

HCM

VUL9

2

HCM

VUL8

3

100.0090.0080.0070.0060.0050.0040.0030.0020.0010.00

0.00

Fo

ld in

cre

ase

NP

C

HC

MV

/NP

C m

ock






nonmalignant, were also induced to produce VEGF, suggest-ing that US28 expression may promote an angiogenicphenotype in normal adult neural cells. An HUVEC tubeformation assay was used to quantify angiogenesis. Figure4C shows that NPC HCMV–infected or overexpressing US28produced supernatant enriched in proangiogenic growth

factors that induced a dramatic increase in HUVEC tubeformation compared with mock infection or transductionwith control vector (Fig. 4D). These data indicate that US28expression in a normal neural precursor cell could stimulateangiogenesis of neighboring endothelial cells. To show spec-ificity of the US28 proangiogenic activity, we carried out

A

B

C

LXSN-US28

60

50

40

30

20

10

0

–10

% c

hange in tre

ate

d/m

ock

HCMV

P3

8

ER

K1

/2

ME

K1

/2

STA

T2

STA

T3

STA

T5

B

STA

T 6

FA

K

SR

C

FY

N

YE

S

FG

R

CR

EB

eN

OS

STA

T5

A/B

AK

T (

S4

73

)

MS

K

D

Figure 3. US28 induces activation ofcellular kinases involved in gliomapathogenesis. A and B, HCMV(Towne;MOI¼ 1) andmock-treatedNPCs (A) and glioma cells (B) wereprofiled with a phosphor-kinasehuman antibody array. C,densitometry measurements weredone per the manufacturer'sinstructions. Percentage of changein phosphorylation levels betweenHCMV/US28-treated and controlcells is shown. One (of 2)representative experiment isshown. D, doubleimmunofluorescence for US28 andthe indicated proteins in NPCstransduced with LXSN-US28 for 48hours. Right, IgG staining controls.Nuclei were counterstained withpropidium iodide. Bar, 50 mm.

Soroceanu et al.





Figure 4. US28 promotes gliomaangiogenesis. A, NPCs transducedwith either LXSN-HA-US28 or Ad-US28 and control LXSN/mock-treated cells were processed forimmunofluorescence. Right, NPCsthat express US28 (green) secreteVEGF (blue), as shown bycolocalization of the 2 markers.Nuclei are stained with propidiumiodide. Bar, 100 mm. B, NPC, U251,U87, and a primary GBM line (4121)were treated with HCMV (Towneand TR; MOI ¼ 1), transduced withAd-US28, or treated with EGF(50 ng/mL) in serum-free media.Supernatants were used in anELISA for VEGF. Samples wereassayed in quadruplicate, and theexperiment was repeated twice.Comparisons between treated andmock within the same cell line wereanalyzed by ANOVA. �, P ¼ 0.02;��, P < 0.002. C, NPC-derivedsupernatants were tested inHUVEC tube formation assays.Complete endothelial cell growthmedia was used as a positivecontrol. Representativephotomicrographsare shown. Eachcondition was assayed in 6 wells ofa 24-well plate, and the experimentwas repeated twice. Bar, 100mm.D,average numbers of branch pointsand endothelial cell lumens areshown from 1 representativeexperiment. Comparisons wereanalyzed by ANOVA. ��, P < 0.02 inall cases.

A

B

VE

GF

(pg/m

L)

Ave

rage/w

ell

Branch points Lumens

NPCU251

GBM# 4121U87

LX

SN TR

US

28

Com

ple

te m

edia

C

D

450

400

350

300

250

200

150

100

50

0

9080706050403020100

Mock

Towne

TR

Ad-U

S28EG

F

**

**

**

** **

** **

**

**

****

****

**** ****

**

*

*






loss-of-function experiments, using siRNA to knock downUS28 in persistently infected glioma lines. US28 knockdowninhibited VEGF production and glioma cell–mediated angio-genesis as measured by HUVEC tube formationassays. Figure 5A illustrates US28 and VEGF detection inpersistently infected U87 glioma cells before and after US28knockdown. We used quantification of immunofluorescencesignals to measure the extent of US28 protein knockdown(Fig. 5B and Supplementary Fig. S5). Cumulative distributionof pixel intensity for immunopositivity illustrates thatapproximately 80% of US28-positive cells lost their signalafter treatment with targeting US28 siRNA1þ2, confirmingeffective protein knockdown (Fig. 5B). A similar level of US28knockdown was achieved in the 4121-HCMV–infected cellsfollowing 72-hour treatment with targeting siRNA1þ2 (datanot shown). VEGF secretion was inhibited by US28 knock-down (Fig. 5B, bottom). Using ELISA, we determined thatVEGF levels (initially induced by HCMV) were inhibited by35% in HCMV-infected U87 and primary glioma cells, fol-lowing US28 knockdown using siRNA1þ2 (Fig. 5C). EachUS28 siRNA used separately had a more modest effect in

inhibiting VEGF secretion, whereas uninfected glioma cellsdid not show a change in VEGF levels, confirming specificityof the US28 knockdown effect (Fig. 5C). Supernatants frompersistently infected glioma cells with or without US28siRNA1þ2 were used in an HUVEC tube formation assay.As shown in Fig. 5D and E, US28 knockdown significantlyinhibited the proangiogenic activities of the HCMV-positiveglioma cell supernatants. US28 knockdown in an endoge-nously infected primary GBM-derived culture inhibitedVEGF secretion by approximately 50% (Supplementary Fig.S6), suggesting potential therapeutic benefits for targetingUS28 in GBM patients.

Further analysis of primary GBM cells from patients iden-tified several tumor cases in which US28 expression wassignificant and where VEGF expression had a high level ofcolocalization with US28 (Fig. 6A–C and Supplementary TableS1). Immunofluorescence analysis of primary GBM cells foreNOS andUS28 indicated that US28 also colocalizedwith eNOS(Fig. 6D–F). Using paraffin-embedded tissue samples from thesame patient, we found HMCV US28, VEGF, e-NOS, and COX2coexpressed both in tumor cells and within the tumor

AUS28/DAPI VEGF/DAPI

Co

ntr

ol

siR

NA

US

28 s

iRN

A 1

+2

VE

GF

(p

g/m

L)

Avera

ge b

ran

ch

es/lu

men

s

B

Perc

enta

ge c

ells

100.00

75.00

50.00

25.00

0.00

1,000

900

800

700

600

500

400

300

200

100

0

60

50

40

30

20

10

0

100.00

75.00

50.00

25.00

0.00Perc

enta

ge c

ells

C

E

D

0 50 100 150

Fluorescence intensity

Co siRNA

US28 expression

VEGF expression

Uninfected HCMV

US28 siRNA 1+2

Co siRNA

Co

siRNA

Neg

ative

cont

rol

Moc

k +

US28

siR

NA1+

2

HCM

V + U

S28 siR

NA1+

2

Posot

ive

cont

rol

Moc

k

HCM

V

US28

siR

NA1

US28

siR

NA2

US28

siR

NA1+

2

Co

siRNA

US28

siR

NA1

US28

siR

NA2

US28

siR

NA1+

2

U87GBM 4121

US28 siRNA 1+2

200 250

0 50 100 150

Fluorescence intensity

**

**

* *

* *

200 250

Figure 5. US28 knockdown inHCMV-infected gliomacells inhibitsVEGF secretion and subsequentangiogenesis. A,immunofluorescence was used fordetectionofUS28 (green) andVEGF(red) in U87 cells persistentlyinfected with HCMV treated eitherwith control siRNA (top) orsiRNA1þ2 targeting US28. Nucleiwere counterstained with DAPI.Bar, 50 mm. B, cumulativedistribution of mean pixel intensityper cell obtained fromimmunofluorescence detection ofUS28 and VEGF in U87 cells treatedwith either targeting (siRNA1þ2) orcontrol siRNAs. The Kolmogorov–Smirnov test was used to determinesignificance of differences in thefluorescence intensity measured inmore than 100 cells per condition,P ¼ 0.0001. C, VEGF levels weremeasured by ELISA in U87 gliomacells and primary 4121 GBM cellsuninfected or HCMV-infected in thepresence of either control or US28targeting siRNA1, siRNA2, orsiRNA1þ2. Differences weresignificant. �, P ¼ 0.05;��, P ¼ 0.002, the Student t test. D,quantification of HUVEC branchesand lumens formed in each of theindicated conditions. ��, P < 0.02,ANOVA. E, representativephotomicrographs of HUVEC tubeformation assay in the presence ofvarious types of conditionedmedia,as indicated. Bar, 100 mm. HUVECtube formation assays wererepeated 3 times, each conditionwas run in quadruplicate.

Soroceanu et al.





Figure 6. HCMV US28 colocalizeswith markers of invasiveness andangiogenesis in situ. A–F, primaryglioblastoma-derived cells wereprocessed for immunofluorescencewith antibodies against US28 (Aand D), VEGF (B), and e-NOS (E). Cand F, merged photomicrographsof colocalization of US28 and the 2markers of angiogenesis. Nuclei arecounterstained with propidiumiodide. Bar, 100 mm. G–L,consecutive paraffin sections (5 mmapart) from a glioblastomaspecimen were stained for US28,VEGF, e-NOS, and COX2 anddeveloped with horseradishperoxidase–3,30-diaminobenzidine. Arrows indicatecells positive for several markers inthe same area. Counterstaining,hematoxylin. Bar, 50 mm. M,summary of the autocrine andparacrine signaling pathwaysthrough which US28 promotesGBM growth, invasion, andangiogenesis.

A B C

D E F

G H I

J

M

K L

p-STAT3

NFKB

COX2

CCL5

CCL5 Invasion, growth

IL-6R

IL-6

US28US28 Extracellular

GBM/neural precursor cell

Blocking peptide + US28 US28 VEGF

Control lgG eNOS COX-2

NO-synthesis

VEGF

Endothelial cell migrationAngiogenesis

SurvivalAkt

eNOS PSer 11/7






microenvironment (Fig. 6G–L), suggesting that proinflamma-tory and proangiogenic signaling is, at least in part, initiatedand promoted by US28 expression in infected GBM cells.Together with the other already described mechanisms, suchas activation of the IL-6–p-STAT3 pathway (8), and inductionof CCL5 (our data), HCMV US28 emerges as a key regulator ofGBM progression by enhancing tumor cell invasion and angio-genesis (Fig. 6M).

Discussion

Our laboratory first identified and reported the presence ofvarious HCMV proteins in human glioblastoma. In this study,we investigated the role of US28 in driving critical signalingpathways supporting glioma growth such as invasion andangiogenesis.

We carried out a systematic screening of human primaryglioma tissues and controls by immunohistochemical andreverse transcriptase PCR/sequencing, which indicate thatapproximately 60% of human GBMs are US28 positive by oneor more techniques. On the basis of these findings, we hypoth-esized that US28 expression in normal neural precursor cellsmight promote gliomagenesis and that US28 expression inestablished GBM cells could promote tumor angiogenesis andinvasion.

The experimental findings we present here support ourhypothesis. When we infected NPCs with either a laboratoryor clinical HCMV isolate, we detected US28 among the mosthighly expressed HCMV genes. Furthermore, HCMV infectionof NPCs resulted in approximately 40-fold increase in CCL5mRNA levels, which could further enhance US28-mediatedsignaling in an autocrine manner because CCL5 binds andactivates US28 (19). CCL5 overexpression has been previouslyassociated with glioblastoma (20), and we determined that itsexpression levels correlate with poor GBM patient outcome byinterrogating a public database.

Our data show for the first time the existence of anautocrine signaling loop in HCMV-infected or US28-expres-sing glioma cells that respond to CCL5 stimulation withincreased invasive behavior. Interestingly, this interactionseems to be cell-type specific, as another study has shownthat macrophages expressing US28 migrate toward Fraktalk-ine (another US28 ligand) rather than CCL5 (21). In thecontext of glioma-associated inflammation, US28 may there-fore modulate multiple autocrine and paracrine loops, pro-moting an oncogenic tumor microenvironment. US28 over-expression also induced approximately 3-fold increase inVEGF levels in both NPCs and GBM cells. Together, thesedata indicate that US28 induces a proangiogenic and inva-sive phenotype both in malignant glioma cells and in NPCs.To ascertain specificity of the proinvasive and proangiogenicactivities of US28, we carried out loss-of-function experi-ments in persistently infected glioma cells and a primaryglioma culture. We used the U87 glioma cell line bearingwell-defined genomic alterations in conjunction with 2custom-made siRNA duplexes used alone and in combina-tion to asses the specificity of US28 effects. Our data showthat US28 knockdown significantly inhibited GBM cell inva-

sion and secretion of biologically active VEGF in HCMV-positive gliomas. These results also suggest that targetingUS28 may have therapeutic benefits for GBM patients.

Immunohistochemical analysis of biopsy specimens ofpatients with glioma showing colocalization of US28 withVEGF, COX2, p-STAT3, and e-NOS in situ suggests once againthat multiple US28-driven mechanisms contributing to theaggressiveness of primaryGBMsmay exist. Our data add to andcorroborate recent reports indicating that HCMVUS28 expres-sion can promote oncogenesis. Maussang and colleagues haveshown that US28 expression can induce oncogenic transfor-mation of 3T3 fibroblasts, and this work was recently extendedupon by the demonstration that a critical mediator or the US28oncogenic signaling pathway is NF-kB, whose activation drivesexpression of VEGF and COX2 (22). Our data indicate thatbiologically active VEGF is induced by US28 in neural pre-cursors and glioma cells and that COX2 and VEGF are coex-pressed with US28 in GBMs.

Slinger and colleagues recently showed thatUS28 expressioncould be detected in human GBM specimens and that itactivates the IL-6–STAT3 signaling pathway in a fibroblastmodel system (8). Our data show for the first time that US28induces p-STAT3 activation in NPCs and documents colocali-zation of US28 and p-STAT3 in primary GBM cultures. A recentreport showed that expression of HCMV US28 in intestinalepithelial cells led to high penetrance of adenocarcinomas in atransgenic mouse model of intestinal neoplasia (23). Thesetumors could be accelerated by coexpression of anUS28 ligand,CCL2.Consistentwith theseobservations, our results showthatUS28 increased glioma cell invasion, which was furtherenhanced by the addition of another US28 ligand, CCL5, aproinflammatory cytokine associated with aggressive GBMs.

Taken together, our data suggest that HCMV US28 may be acritical factor in promoting the transformation ofNPCs and theinvasive and angiogenic properties of established glioblasto-mas. Our results suggest that targeting specificHCMVproteins(e.g., US28) in endogenously infected GBMs may disruptcritical pathways and constitute a novel antitumor approach.

Disclosure of Potential Conflicts of Interest

The authors declare no potential conflicts of interest.

Acknowledgments

The authors thank Dr. Dan Streblow for the Ad-US28 and control adeno-viruses, Dr. Martine Smit for the pcDEF-US28 expression plasmid, Dr. LeeFortunato for the TR virus, and Dr. Dan Moore for help with the microarraydata analysis. The authors also thank Sabeena Khan for technical help.

Grant Support

These studies were supported by NIH grants R01NS070289-02 to C.S. Cobbs,1R21NS067395-01 to L. Soroceanu, ACS grant RSG-09-197-01 to C.S. Cobbs, andadditional funds from the ABC2 Foundation and the Flaming Foundation.

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 March 1, 2011; revised August 30, 2011; accepted August 31, 2011;published OnlineFirst September 7, 2011.

Soroceanu et al.





References1. Cobbs CS, Harkins L, Samanta M, Gillespie GY, Bharara S, King PH,

et al. Human cytomegalovirus infection and expression in humanmalignant glioma. Cancer Res 2002;62:3347–50.

2. Mitchell DA, Xie W, Schmittling R, Lean C, Friedman A, McLendon RE,et al. Sensitive detection of human cytomegalovirus in tumors andperipheral bloodof patientsdiagnosedwithglioblastoma.NeuroOncol2008;10:10–8.

3. Soroceanu L, Cobbs CS. Is HCMV a tumor promoter? Virus Res2011;157:193–3.

4. Cobbs CS, Soroceanu L, Denham S, ZhangW, Kraus MH. Modulationof oncogenic phenotype in human glioma cells by cytomegalovirusIE1-mediated mitogenicity. Cancer Res 2008;68:724–30.

5. Soroceanu L, Akhavan A, Cobbs CS. Platelet-derived growth factor-alpha receptor activation is required for human cytomegalovirus infec-tion. Nature 2008;455:391–5.

6. Vischer HF, Leurs R, Smit MJ. HCMV-encoded G-protein-coupledreceptors as constitutively active modulators of cellular signalingnetworks. Trends Pharmacol Sci 2006;27:56–63.

7. Maussang D, Verzijl D, van Walsum M, Leurs L, Holl J, Pleskoff O,et al. Human cytomegalovirus-encoded chemokine receptor US28promotes tumorigenesis. Proc Natl Acad Sci U S A 2006;103:13068–73.

8. Slinger E, Maussang D, Schreiber A, Siderius M, Rahbar A, Fraile-Ramons A, et al. HCMV-encoded chemokine receptor US28mediatesproliferative signaling through the IL-6-STAT3 axis. Sci Signal 2010;3:ra58.

9. Soroceanu L, Kharbanda S, Chen R, Soriano RH, Aldape K, Misra A,et al. Identification of IGF2 signaling through phosphoinositide-3-kinase regulatory subunit 3 as a growth-promoting axis in glioblasto-ma. Proc Natl Acad Sci U S A 2007;104:3466–71.

10. CobbsCS, Soroceanu L, DenhamS, ZhangW,BrittWJ, Pieper R, et al.Human cytomegalovirus induces cellular tyrosine kinase signaling andpromotes glioma cell invasiveness. J Neurooncol 2007;85:271–80.

11. YangS,GhannyS,WangW,Galante A,DunnW, Liu F, et al. UsingDNAmicroarray to study human cytomegalovirus gene expression. J VirolMethods 2006;131:202–8.

12. Phillips HS, Kharbanda S, Chen R, Forest WF, Soriano RH, Wu TD,et al. Molecular subclasses of high-grade glioma predict prognosis,

delineate a pattern of disease progression, and resemble stages inneurogenesis. Cancer Cell 2006;9:157–73.

13. Network TCGAR. Comprehensive genomic characterization defineshuman glioblastoma genes and core pathways. Nature 2008;455:1061–8.

14. CobbsCS,BrenmanJE,AldapeKD,BredtDS, IsraelMA.Expression ofnitric oxide synthase in human central nervous system tumors. CancerRes 1995;55:727–30.

15. Cobbs CS, Fenoy A, Bredt DS, Noble LJ. Expression of nitric oxidesynthase in the cerebralmicrovasculature after traumatic brain injury inthe rat. Brain Res 1997;751:336–8.

16. Charles N, Holland EC. The perivascular niche microenvironment inbrain tumor progression. Cell Cycle 2010;9:3012–21.

17. Carro MS, LimWK, Alvarez MJ, Bolo RJ, Zhao X, Snyder EP, et al. Thetranscriptional network for mesenchymal transformation of braintumours. Nature 2010;46:318–25.

18. Dziurzynski K, Wei J, Qiao W, Hatiboglu MA, Kong LY, Wu A, et al.Glioma-associated cytomegalovirus mediates subversion of themonocyte lineage to a tumor propagating phenotype. Clin CancerRes 2011;17:4642–9.

19. Streblow DN, Soderberg-Naucler C, Vieira J, Smith P,Wakabayashi E,Ruchti F, et al. The human cytomegalovirus chemokine receptor US28mediates vascular smoothmuscle cellmigration.Cell 1999;99:511–20.

20. Maru SV, Holloway KA, FlynnG, Lancashire CL, Loughlin AJ, Male DK,et al. Chemokine production and chemokine receptor expression byhuman glioma cells: role of CXCL10 in tumour cell proliferation. JNeuroimmunol 2008;199:35–45.

21. VomaskeJ,Nelson JA,StreblowDN.HumanCytomegalovirusUS28: afunctionally selective chemokine binding receptor. Infect Disord DrugTargets 2009;9:548–56.

22. Maussang D, Langemeijer E, Fitzsimons CP, Stigter-van Walsum M,Dijkman R, Borg MK, et al. The human cytomegalovirus-encodedchemokine receptor US28 promotes angiogenesis and tumor forma-tion via cyclooxygenase-2. Cancer Res 2009;69:2861–9.

23. Bongers G, Maussang D, Muniz LR, Noriega VM, Fraile-Ramos A,Barker N, et al. The cytomegalovirus-encoded chemokine receptorUS28 promotes intestinal neoplasia in transgenic mice. J Clin Invest2010;120:3969–78.






2011;71:6643-6653. Published OnlineFirst September 7, 2011.Cancer Res Liliana Soroceanu, Lisa Matlaf, Vladimir Bezrookove, et al. Invasive and Angiogenic PhenotypeHuman Cytomegalovirus US28 Found in Glioblastoma Promotes an

Updated version

10.1158/0008-5472.CAN-11-0744doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2011/10/21/0008-5472.CAN-11-0744.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/71/21/6643.full#ref-list-1

This article cites 23 articles, 8 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/71/21/6643.full#related-urls

This article has been cited by 10 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/71/21/6643To 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-11-0744

http://cancerres.aacrjournals.org/content/suppl/2011/10/21/0008-5472.CAN-11-0744.DC1

http://cancerres.aacrjournals.org/content/71/21/6643.full#ref-list-1

http://cancerres.aacrjournals.org/content/71/21/6643.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/71/21/6643


Human Cytomegalovirus US28 Found in Glioblastoma Promotes ... › content › canres › ... ·...

Documents

Transcript of Human Cytomegalovirus US28 Found in Glioblastoma Promotes ... › content › canres › ... ·...