of December 19, 2018.This information is current as

Tumor Suppression In VivoLigand-Induced Apoptosis In Vitro andSensitize Colon Carcinoma Cells to Fas Decitabine and Vorinostat Cooperate To

Zimmerman, Tracy L. McGaha and Kebin LiuDafeng Yang, Christina M. Torres, Kankana Bardhan, Mary

ol.1103035http://www.jimmunol.org/content/early/2012/03/28/jimmun

published online 28 March 2012J Immunol 

MaterialSupplementary

5.DC1http://www.jimmunol.org/content/suppl/2012/03/28/jimmunol.110303

        average*  

4 weeks from acceptance to publicationFast Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

Submit online. ?The JIWhy

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists, Inc. All rights reserved.Copyright © 2012 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on Decem

ber 19, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

by guest on D

ecember 19, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

The Journal of Immunology

Decitabine and Vorinostat Cooperate To Sensitize ColonCarcinoma Cells to Fas Ligand-Induced Apoptosis In Vitroand Tumor Suppression In Vivo

Dafeng Yang,* Christina M. Torres,* Kankana Bardhan,* Mary Zimmerman,*

Tracy L. McGaha,†,‡,x and Kebin Liu*,‡

The death receptor Fas and its physiological ligand (FasL) regulate apoptosis of cancerous cells, thereby functioning as a critical

component of the host cancer immunosurveillance system. To evade Fas-mediated apoptosis, cancer cells often downregulate Fas

to acquire an apoptosis-resistant phenotype, which is a hallmark of metastatic human colorectal cancer. Therefore, targeting Fas

resistance is of critical importance in Fas-based cancer therapy and immunotherapy. In this study, we demonstrated that

epigenetic inhibitors decitabine and vorinostat cooperate to upregulate Fas expression in metastatic human colon carcinoma

cells. Decitabine also upregulates BNIP3 and Bik expression, whereas vorinostat decreased Bcl-xL expression. Altered expression

of Fas, BNIP3, Bik, and Bcl-xL resulted in effective sensitization of the metastatic human colon carcinoma cells to FasL-induced

apoptosis. Using an experimental metastasis mouse model, we further demonstrated that decitabine and vorinostat cooperate to

suppress colon carcinoma metastasis. Analysis of tumor-bearing lung tissues revealed that a large portion of tumor-infiltrating

CD8+ T cells are FasL+, and decitabine and vorinostat-mediated tumor-suppression efficacy was significantly decreased in Fasgld

mice compared with wild-type mice, suggesting a critical role for FasL in decitabine and vorinostat-mediated tumor suppression

in vivo. Consistent with their function in apoptosis sensitization, decitabine and vorinostat significantly increased the efficacy of

CTL adoptive transfer immunotherapy in an experimental metastasis mouse model. Thus, our data suggest that combined

modalities of chemotherapy to sensitize the tumor cell to Fas-mediated apoptosis and CTL immunotherapy is an effective

approach for the suppression of colon cancer metastasis. The Journal of Immunology, 2012, 188: 000–000.

The Fas-mediated apoptosis pathway was originally iden-tified to play a critical role in the immune system fordepletion of self-reactive lymphocytes. Germline and so-

matic mutations or deletions of Fas- or its physiological ligand(FasL)-coding sequences in humans lead to autoimmune lym-phoproliferative syndrome (1–5). Mice that are deficient in Fasor FasL also develop lymphoproliferation disorder, resulting inlymphadenopathy and systemic lupus-like autoimmune disease (6,7). These observations indicate that Fas plays a critical role inimmune cell homeostasis and in the suppression of autoimmunediseases. However, it has become increasingly appreciated that theFas-mediated apoptosis pathway is also directly involved insuppression of tumor development (8–12). Human autoimmunelymphoproliferative syndrome patients exhibited an increased risk

for both hematopoietic and nonhematopoietic cancers (1, 4, 13).Furthermore, both the Fas and FasL gene promoters are poly-morphic, including a G-to-A substitution at 21377 bp and an A-to-G substitution at 2670 bp in the Fas gene promoter, as well asa C-to-T substitution at 2844 bp and an A-to-G substitution at2124 bp in the FasL gene promoter. These polymorphisms di-minish transcription factor binding to the Fas and FasL promoterto decrease Fas and FasL expression levels, resulting in an in-creased risk for cancer development in humans (14–19). More-over, a study with a large cohort of human colorectal cancerpatient specimens showed that Fas-mediated apoptosis is an im-portant contributor to tumor regression (11). Therefore, the Fasand FasL system also regulate apoptosis of cancerous cells and,thus, functions as a critical component of the host cancer immu-nosurveillance system against cancer development (20–22).The efficacy of cytotoxicity-based cancer therapy largely

depends on induction of tumor cell apoptosis. The critical roleof Fas in tumor cell apoptosis makes targeting the Fas-mediatedapoptosis pathway an attractive approach in cancer therapy.FasL protein and anti-Fas agonist Abs are potentially effectiveanticancer agents. However, Fas-based chemotherapies are likelyhighly toxic because infusion of FasL protein or anti-Fas agonistAbs induces extensive apoptosis of hepatocytes, resulting in lethalliver damage (23–25), thereby limiting the clinical use of FasLprotein or anti-Fas Abs for systemic anticancer chemotherapy. Incontrast, FasL is expressed on activated CTLs, and tumor-specificFasL+ CTLs are natural biological agents for inducing Fas-mediated apoptosis in cancer therapy (11, 26). However, cancercells often silence Fas expression and/or acquire an apoptosis-resistant phenotype to evade Fas-mediated killing. For example,Fas is constitutively expressed at high levels in normal human

*Department of Biochemistry and Molecular Biology, Georgia Health Sciences Uni-versity, Augusta, GA 30912; †Department of Medicine, Georgia Health SciencesUniversity, Augusta, GA 30912; ‡Cancer Center, Georgia Health Sciences University,Augusta, GA 30912; and xImmunotherapy Center, Georgia Health Sciences Univer-sity, Augusta, GA 30912

Received for publication October 21, 2011. Accepted for publication February 25,2012.

This work was supported by the National Institutes of Health (CA133085 to K.L.)and the American Cancer Society (RSG-09-209-01-TBG to K.L.).

Address correspondence and reprint requests to Dr. Kebin Liu, Department of Bio-chemistry and Molecular Biology, Georgia Health Sciences University, 1410 LaneyWalker Boulevard, Augusta, GA 30912. E-mail address: Kliu@georgiahealth.edu

The online version of this article contains supplemental material.

Abbreviations used in this article: FasL, Fas ligand; HDAC, histone deacetylase; MS,methylation sensitive; MTD, maximally tolerated dose; PI, propidium iodide; siRNA,small interfering RNA; Treg, regulatory T cell; wt, wild-type.

Copyright� 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1103035

Published March 28, 2012, doi:10.4049/jimmunol.1103035 by guest on D

ecember 19, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

colon tissues; however, in human primary colorectal carcinoma,Fas expression is often diminished, and complete loss of Fas ex-pression is often observed in metastatic human colorectal carci-noma (27, 28). Thus, resistance to Fas-mediated apoptosis isa major obstacle of Fas-based CTL immunotherapy against met-astatic human colorectal cancer.Decitabine is a cytidine analog that inhibits DNA methyl-

transferase activity upon incorporation into replicating DNA and isan approved agent for myelodysplastic syndrome. Decitabine wasinitially used at or near the maximally tolerated dose (MTD), atwhich it has a cytotoxic effect, to treat solid tumors, but was foundto be associated with severe toxicity and minimal efficacy (29). Itwas later observed that decitabine, at a dose well below its MTD,is effective in the inhibition of DNA methylation and achieveslong-term tolerance and improved clinical efficacy in patientswith myelodysplastic syndrome and solid tumors (30, 31). Vor-inostat is a histone deacetylase (HDAC) inhibitor and is an ap-proved agent for the treatment of cutaneous T cell lymphoma.Vorinostat, at or near its MTD, is also associated with severetoxicity and exhibits minimal efficacy in solid tumors when usedas a single agent (32). Although the expression of many genesis modulated by vorinostat, vorinostat alone is often ineffectivein the induction of expression of hypermethylated genes (33).Therefore, vorinostat and decitabine are often combined to ach-ieve maximal activation efficacy of epigenetically silenced genesin cancer cells (31, 33–35).Pioneer studies demonstrated that decitabine and vorinostat can

overcome apoptosis resistance in various types of cancers (35–37).These epigenetic inhibitors were shown to either reactivate theexpression of death receptor Fas in tumor cells (36–38) or targetthe Fas-mediated apoptosis-signaling pathways to induce tumorcell apoptosis (33–35, 39, 40). Based on these observations, wehypothesized that epigenetic mechanism-based chemotherapymay be combined with CTL immunotherapy to overcome tumorcell Fas resistance to increase the efficacy of CTL immunotherapy.This idea is analogous to a “one-two punch” strategy. First, can-cer cells are treated with apoptosis-sensitizing drugs to activateFas and/or sensitize tumor cells to Fas-mediated apoptosis. Once“sensitized,” tumors are treated with FasL+ tumor-specific CTLsthat promote Fas-mediated apoptosis to destroy the tumors. Totest this hypothesis, we performed this proof-of-concept studyand identified that epigenetic inhibitors decitabine and vorinostatregulate the expression of Fas, BNIP3, Bik, and Bcl-xL to coop-eratively sensitize the metastatic human colon carcinoma cells toFasL-induced apoptosis. Furthermore, we demonstrated that dec-itabine and vorinostat-mediated tumor suppression depends, atleast in part, on FasL in vivo. Overall, our results indicate thatcombined Fas-based chemotherapy and FasL-dependent CTLimmunotherapy is effective in suppressing colon carcinoma me-tastasis, and it holds great promise for further development for thetreatment of metastatic human colorectal cancer.

Materials and MethodsMice

Fasgld (CPt.C3-Faslgld/J), BALB/cByJ mice were obtained from TheJackson Laboratory. BALB/c mice were obtained from the NationalCancer Institute (Frederick, MD). All mice were used at age $6 wk.Experiments and care/welfare were in agreement with federal regulationsand an approved protocol by the Georgia Health Sciences UniversityAnimal Care and Use Committee.

Reagents

Decitabine was obtained from Sigma (St. Louis, MO). Vorinostat (Merck)was provided by the Cancer Treatment and Evaluation Program, National

Cancer Institute/National Institutes of Health. FasL (Mega-Fas Ligand,kindly provided by Drs. Steven Butcher and Lars Damstrup, Topotarget,Copenhagen, Denmark) is a recombinant fusion protein that consists ofthree human FasL extracellular domains linked to a protein backbonecomprising the dimer-forming collagen domain of human adiponectin. TheMega-Fas Ligand was produced as a glycoprotein in mammalian cellsusing a Good Manufacturing Practice-compliant process at Topotarget.

RT-PCR analysis

Total RNA was isolated from cells or tissues using TRIzol reagent (Invi-trogen, San Diego, CA) and used for real-time RT-PCR analysis of geneexpression, as described (41, 42). The PCR primer sequences are as fol-lows: human Fas: forward: 59-ATTATCGTCCAAAAGTGTTAAT-39, re-verse: 59-TGCATGTTTTCTGTACTTCCTT-39; mouse FasL: forward: 59-CTTGGGCTCCTCCAGGGTCAGT-39, reverse: 59-TCTCCTCCATTAG-CACCAGATCC-39; and b-actin: forward: 59-ATTGTTACCAACTGGG-ACGACATG-39, reverse: 59-CTTCATGAGGTAGTCTGTCAGGTC-39.

Cell treatments

For treatment with decitabine, cells were cultured in its presence for 3 d. Forvorinostat treatment, cells were cultured in its presence for 2 d

Sodium bisulfite treatment and DNA-methylation analysis

Genomic DNAwas isolated using a DNeasy Tissue Kit (QIAGEN). Sodiumbisulfite treatment of genomic DNA was carried out using a CpGenomeUniversal DNA Modification Kit (Chemicon, Temecula, CA). Methylation-sensitive (MS)-PCRwas carried out as previously described (43). The primersequences are as follows: BNIP3: unmethylation: forward: 59-TGTTTTT-TTAAAGGAGAATTTGG-39, reverse: 59-CAAAAAACAAAAACCTACA-ATACAC-39, methylation: forward: 59-TTATCGTTTTTTTAAAGGAGA-ATTC-39, reverse: 59-GAAAAACAAAAACCTACGATACG-39 and Bik:unmethylation: forward: 59-GTAATGGGATGTTTAGAAAGTTTGG-39, re-verse: 59-AAAAAACCAAAACCCAAACAACAAC-39, methylation: for-ward: 59-GTAACGGGACGTTTAGAAAGTTCGG-39, reverse: 59-TAAA-AAACCAAAACCCAAACGACG-39. The bisulfite-modified genomicDNA was also used as a template for PCR amplification of the modifiedgenomic DNA fragments of the human Fas promoter regions. The PCRprimers are as follows: region 1: forward: 59-GGGATAGGAATGTTTAT-TTGTGTAA-39, reverse: 59-CCTAAAACTCCAACCAAATCACT-39 andregion 2: forward: 59-GGAGGATTGTTTAATAATTATGTTGG-39, re-verse: 59-TCTAAAAACTACAAACTCTCTCCCC-39. The PCR fragmentswere cloned to pCR2.1 vector (Invitrogen), and plasmid was purified fromindividual clones and sequenced.

Western blotting analysis

Western blotting analysis was performed as previously described (44). Thefollowing primary Abs were obtained from Cell Signaling Biotech (Dan-vers, MA): Bak, Bik, Bid, FLIP, cIAP1, xIAP, Bad, Bok, and PUMA. Thefollowing primary Abs were obtained from Santa Cruz Biotechnology(Santa Cruz, CA): Bax, survivin, Mcl-1, and BNIP3. The following Abswere obtained from BD Biosciences: Bcl-2 and Bcl-xL. Anti–b-actin wasobtained from Sigma (St. Louis, MO).

Apoptosis assays

Cells were either stained with propidium iodide (PI) (Trevigen, Gaithers-burg, MD) or PI plus Annexin V-Alexa Fluor 647 (BioLegend, San Diego,CA) and analyzed by flow cytometry.

Cell surface protein analysis

Tumor cells were stained with anti-Fas (BD Biosciences) mAb. Isotype-matched control IgG (BD Biosciences) was used as a negative control.The stained cells were analyzed by flow cytometry. For FasL proteinanalysis, mouse lungs were digested in collagenase solution to create asingle-cell suspension. The cell suspension was stained with PE-conjugatedFasL (BD Biosciences), FITC-conjugated CD8 mAb (BioLegend), or bothmAbs and analyzed by flow cytometry.

Gene silencing

Tumor cells were transiently transfected with scramble small interferingRNA (siRNA; Dharmacon, Lafayette, CO) and human Bcl-xL–specificsiRNA (Santa Cruz Biotechnology), respectively. Cells were then har-vested and cultured in 24-well plates overnight in the absence or presenceof FasL before analysis for apoptosis.

2 TARGETING Fas RESISTANCE TO SUPPRESS COLON CANCER

by guest on Decem

ber 19, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

Tumor cell transfection

Tumor cells were transiently transfected with the pEGFP control vector andpEGFP.hBik plasmid, respectively. Cells were then harvested and culturedovernight in 24-well plates in the absence or presence of FasL beforeanalysis for apoptosis.

Liver-toxicity analysis

Decitabine (0.1 mg/kg body weight) and vorinostat (25 mg/kg body weight)were injected i.v. into BALB/c mice either alone or in combination. Serum

was collected from mice 3 d later and measured for complete liver profile

(Table I) at Georgia Laboratory Animal Diagnostic Service (Athens, GA).

Experimental lung metastasis mouse model and CTLimmunotherapy

Tumor-specific CTLs were generated from perforin-deficient BALB/cmice, as previously described (45). The experimental lung metastasis

mouse model and CTL adoptive-transfer immunotherapy were carried

out as previously described (45). Decitabine was used at a dose of

Fas Fas

Cell

counts

Cell

counts

Control +Decitabine

+Vorinostat +Decitabine

& Vorinostat

02468

10

Fas M

FI

Control Decitabine Vorinostat Decitabine

&Vorinostat

p<0.01p<0.01

A B

C

D

0 μM 0.5 μM0.1 μM

1 μM0.75 μM

Fas MFI

Fas MFI

Fas MFI

Cell

counts

Cell

counts

0 μM 0.5 μM0.1 μM

1 μM0.75 μM

Fas MFI

Fas MFI

Fas MFI

Cell

counts

Cell

counts

p=0.04

p<0.01

%F

asL

-in

du

ced

ce

ll d

ea

th

0

20

40

60

Control +Decitabine +Vorinostat

PI PI PI PI

SS

C-H

SS

C-H

+Decitabine

&Vorinostat9.4%

16.7%

17.4%

54.7%

29.3%

44.5%

25%

78.8%

-FasL

+FasL

Fas M

FI

0

2

4

6

8

0 0.1 0.5 0.75 1.0

Decitabine (μM)

Fas M

FI

0

2

4

6

0 0.1 0.5 0.75 1.0

Vorinostat (μM)

Control Decitabine Vorinostat Decitabine

&Vorinostat

Control Decitabine Vorinostat Decitabine

&Vorinostat

Rela

tive

Fas m

RN

Ale

vel

0

1

2

3 p=0.05

p=0.05

p<0.01

p=0.01p<0.01

p<0.01

p<0.01

p=0.2 p=0.02

p<0.01

p=0.02

FIGURE 1. Decitabine and vorinostat cooperate to upregulate Fas expression and sensitize the metastatic human colon carcinoma cells to FasL-induced

apoptosis. LS411N cells were treated with decitabine for 3 d (A) or vorinostat for 2 d (B) at the indicated doses, stained with Fas-specific mAb, and

analyzed by flow cytometry for cell surface Fas protein level. Gray area: IgG isotype-control staining; solid line: Fas-specific staining. Lower panels, The

Fas protein level was quantified as mean fluorescence intensity (MFI). Column: mean; bar: SD. (C) LS411N cells were treated with decitabine (0.5 mM, 3

d), vorinostat (0.5 mM, 2 d), or both decitabine and vorinostat and analyzed for Fas protein level by flow cytometry (left and middle panels). The Fas protein

level is quantified as MFI (middle panel). Column: mean; bar: SD. Fas mRNA level was measured by real-time RT-PCR (right panel). The Fas mRNA level

in control cells was arbitrarily set at 1. (D) Tumor cells were treated as in (C), followed by incubation with FasL (200 ng/ml) overnight, staining with PI, and

analysis by flow cytometry (left panels). The percentage of FasL-induced cell death was quantified as percentage of PI+ cells in the presence of FasL 2percentage of PI+ cells in the absence of FasL (right panel). Column: mean; bar: SD.

The Journal of Immunology 3

by guest on Decem

ber 19, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

0.1 mg/kg body weight, and vorinostat was used at a dose of 20 mg/kgbody weight.

Statistical analysis

Where indicated, data are represented as the mean 6 SD. Statisticalanalysis was performed using a two-sided t test, with p values , 0.05considered statistically significant.

ResultsDecitabine and vorinostat cooperate to upregulate Fasexpression and sensitize metastatic human colon carcinomacells to FasL-induced apoptosis

It was shown that vorinostat activates Fas gene expression in tumorcells (36, 37, 46), whereas Fas promoter DNA methylation hasbeen observed in certain colon carcinoma cells (47). Based onthese observations, we reasoned that inhibition of DNA methyl-ation and HDAC activity may upregulate Fas expression in met-astatic human colon carcinoma cells. To test this notion, themetastatic human colon carcinoma cell line LS411N was treatedwith decitabine and vorinostat and analyzed for Fas expression.Both decitabine and vorinostat increased Fas protein level on thetumor cell surface in a dose-dependent manner, and the increasereached a plateau at a dose of ∼0.75 mM (Fig. 1A, 1B). Inter-estingly, combined decitabine and vorinostat treatment resulted ina significantly higher level of Fas protein than did either agentalone (Fig. 1C). Decitabine and vorinostat also increased the FasmRNA level, but combined decitabine and vorinostat did notfurther increase the Fas mRNA level compared with either agentalone (Fig. 1C).Metastatic human colon carcinoma LS411N cells are highly

resistant to FasL-induced apoptosis. To determine whether theincreased Fas expression leads to increased sensitivity of the tumorcells to FasL-induced apoptosis, LS411N cells were treated withdecitabine and vorinostat, either alone or in combination, and thenincubated with FasL protein. Analysis of cell death revealed thatdecitabine or vorinostat treatment alone increased the tumor cellsensitivity to FasL-induced apoptosis (Fig. 1D). However, con-sistent with the Fas protein level, combined treatment with the two

agents rendered the metastatic human colon carcinoma cells moresensitive to FasL-induced apoptosis than did either treatmentalone (Fig. 1D). Taken together, our data suggest that inhibition ofboth DNA methylation and HDAC activity is an effective ap-proach to overcome apoptosis resistance in metastatic humancolon carcinoma cells.

Fas promoter DNA is sporadically methylated in metastatichuman colon carcinoma cells

To determine whether decitabine upregulates Fas expressionthrough inhibition of the Fas promoter DNA methylation, we ana-lyzed the Fas promoter DNA methylation status in the metastatichuman colon carcinoma cell lines LS411N and SW620. Analysisof the human Fas gene revealed that the human Fas gene promotercontains multiple classical CpG islands surrounding the tran-scription-initiation site (Fig. 2A). However, analysis of the ge-nomic DNA sequence in two regions of the Fas promoter in-dicated that the Fas promoter is not methylated in LS411N cells(Fig. 2B). In SW620 cells, we observed that only 1–3 cyto-sines of the 34 CpGs analyzed are methylated (Fig. 2B).Therefore, we conclude that Fas upregulation by decitabine is un-likely to occur through inhibition of Fas promoter DNA methy-lation.

-63 +208 +328 +609

LS411N

-10 +189 +397 +547 +592 +742 +762 +895 Island 1 Island 2 Island 3 Island 4CpG Island:

SW620

+1 +346

5’-UTR

Transcription initiation

% G

C

+1000-50010080

60

4020

0CpG

Fas PromoterA

B

FIGURE 2. The Fas promoter DNA methylation status in metastatic

human colon carcinoma cells. Upper panel, The human Fas gene pro-

moter, showing CpG islands (gray areas). The vertical bars under the line

indicate the locations of CpG dinucleotides. The locations of the CpG

islands relative to the Fas transcription initiation site (+1) are also in-

dicated under the line. Lower panel, Methylation status of the human Fas

gene promoter in LS411N and SW620 cells. The indicated regions of

bisulfite-modified genomic DNA isolated from LS411N and SW620

cells were cloned and sequenced. s, Unmethylated CpG; d, methylated

CpG.

-Bad

-Bak

-Bax

-Bok

-PUMAα

-PUMAβ

-Bid

-β-actin

-Bcl2

-cIAP1

-cIAP2

-FLIPL

-Mcl-1

-Survivin

-xIAP

-β-actin

-Bcl-xL

-Bik

-BNIP3

-β-actin

Con

trol

Dec

itabine

C

Vorin

osta

t

Dec

itabine

& Vor

inos

tat

Con

trol

Dec

itabine

Vorin

osta

t

Dec

itabine

& Vor

inos

tat

Con

trol

Dec

itabine

Vorin

osta

t

Dec

itabine

& Vor

inos

tat

U MU MLS

411N

SW62

0

U MU MLS

411N

SW62

0

A

B

C

% G

C

100

806040

200

CpG

-2000 -1000 +1 +1000

Transcription

BNIP3 Promoter

100

80

6040

200

CpG

% G

C

-2000 -1000 +1 +1000

Bik PromoterTranscription

FIGURE 3. Decitabine and vorinostat alter the level of apoptosis-related

proteins. (A) Western blotting analysis of both pro- and antiapoptosis

proteins. LS411N cells were treated with decitabine, vorinostat, or both

agents as in Fig. 1 and analyzed for the indicated proteins. Methylation

status of BNIP3 (B) and Bik (C) promoters in colon carcinoma cells.

BNIP3 (B) and Bik (C) promoter structure (left panels). The CpG islands

are indicated by the gray-filled areas. Methylation status of BNIP3 (B) and

Bik (C) promoter DNA (right panels). Bisulfite-modified genomic DNA

was analyzed by MS-PCR to detect DNA methylation. M, Methylated; U,

unmethylated.

4 TARGETING Fas RESISTANCE TO SUPPRESS COLON CANCER

by guest on Decem

ber 19, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

Decitabine and vorinostat regulate the expression of BNIP3,Bik, and Bcl-xL

Sensitivity to Fas-mediated apoptosis is mediated at both the deathreceptor Fas level and within its downstream signaling pathway

(48–50). Therefore, we next analyzed the key mediators of the

Fas-signaling pathway in decitabine and vorinostat-treated met-

astatic human colon carcinoma cells. Western blotting analysis

revealed that protein levels of BNIP3 and Bik were increased

Annexin

VA

nnexin

V

Annexin

VA

nnexin

V

PI PI PI PI

-FasL

+FasL

Scramble Bcl-xsiRNA pEGFP pEGFP.hBik

2.6% 1.8% 5.8% 3.7%

3% 2.7% 24% 14.6%

4.6% 5.4%

3% 5%

4.6% 5.8%

9.5% 11%

0

10

20

30

40

0

5

10

15

LS411N.VectorLS411N.Bik

% A

popto

tic c

ell

death

% A

popto

tic c

ell

death

Scramble

Bcl-x siRNA

Annexin V-

positive

Annexin V &

PI-Positive

Annexin V-

positive

Annexin V &

PI-Positive

C

B

A

D

p<0.01 p<0.01

FIGURE 4. Decitabine and vorinostat sensitize co-

lon carcinoma cells to FasL-induced apoptosis, in part

through altering Bcl-xL and Bik expression. (A) Si-

lencing Bcl-xL increased tumor cell sensitivity to FasL-

induced apoptosis. LS411N cells were transfected with

scramble and Bcl-xL–specific siRNAs for ∼24 h.

Transfected cells were then incubated with FasL (200

ng/ml) overnight, stained with Annexin V and PI, and

analyzed for apoptosis by flow cytometry. Number in

each plot indicates the percentage of PI+ and Annexin

V+ cells. (B) Apoptotic cell death was quantified as

percentage of PI and Annexin V double-positive cells

in the presence of FasL 2 percentage of PI and

Annexin V double-positive cells in the absence of

FasL. Column: mean; bar: SD. (C) Overexpressing Bik

increased tumor cell sensitivity to FasL-induced apo-

ptosis. LS411N cells were transiently transfected with

vector (pEGFP) or Bik-expressing vector (PEGFP.

hBik) and analyzed for sensitivity to FasL-induced apo-

ptosis as in (A). (D) Apoptotic cell death was quantified

as in (B).

Con

trol

Decitabine

&Vorinostat

SS

C-H

SS

C-H

IgG-FITC CD8-FITC FasL-PE

IgG-PE FasL-PE FasL-PE

2.0%

1.9%

7.9%

9.8%

32.3%

7.0%

Dec

itabine

Vorin

osta

t

a b c

d e f

% F

asL

c

ells

+

0

10

20

30

40

CD8 cells-

CD8 cells+

p=0.03

A B

C

D

0

0.5

1

1.5

rela

tive

FasL leve

l

Tumor-free

mice

Tumor-bearing

mice

0

50

100

150

200

250

No. T

um

or

no

du

le

/mo

use

p<0.01

p<0.01

p<0.01

p<0.01

p<0.01

p=0.14

FIGURE 5. Decitabine and vorinostat cooperate to suppress metastatic colon carcinoma growth in vivo. (A) CT26 cells (7.5 3 104 cells/mouse) were

injected into mice. Decitabine (0.1 mg/kg body weight), vorinostat (25 mg/kg body weight), or both decitabine and vorinostat were injected i.v. into tumor-

bearing mice 4 d later. The treatments were repeated every 2 d seven times. Mice were sacrificed 20 d after tumor transplantation and analyzed for lung

metastasis. Images of lungs from representative mice are shown. Lung tumor nodules were enumerated (lower panel). Each circle represents total tumor counts

from a single mouse. Counts.250 are expressed as 250. (B) Real-time RT-PCR analysis of FasLmRNA level in mouse lungs. Lungs were excised from tumor-

free control mice (n = 3) and tumor-bearingmice, as shown in (A) (n = 4) and extracted for total RNA. The FasLmRNA of control mouse 1 was set at 1. (C) Cell

surface FasL protein level. Single-cell suspensions were prepared from lungs of tumor-bearing mice, as shown in (A), stained with CD8- and FasL-specific

mAbs, and analyzed by flow cytometry. IgG isotype control was used as a negative control (a and d). Percentage of CD8+ T cells (b), FasL+CD8+ T cells (c),

FasL+ total lung cells (e), and FasL+CD82 cells (f) were quantified and are indicated in each plot. Shown are representative results from one of five mice. (D)

Quantification of FasL+ cells in CD8+ and CD82 cell populations in tumor-bearing mouse lungs. Data are mean from five mice. Column: mean; bar: SD.

The Journal of Immunology 5

by guest on Decem

ber 19, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

following decitabine treatment, whereas Bcl-xL protein was de-creased by vorinostat treatment (Fig. 3A). Thus, inhibition ofDNA methylation and HDAC activity altered the expressionlevels of multiple apoptosis-related mediators. Analysis of theBNIP3 and Bik promoters revealed that there are classical CpGislands in these two promoter regions around the transcription-initiation sites (Fig. 3B, 3C). MS-PCR analysis indicated thatboth BNIP3 and Bik promoter DNA was hypermethylated inthe metastatic human colon carcinoma cell lines LS411N andSW620 (Fig. 3B, 3C). Thus, our data demonstrated that the pro-apoptotic BNIP3 and Bik gene promoters are silenced by DNAmethylation in metastatic human colon carcinoma cells and thatdecitabine inhibits DNA methylation to reactivate BNIP3 andBik.

Bik, BNIP3, and Bcl-xL mediate apoptosis resistance inmetastatic human colon carcinoma cells

The above observations that decitabine and vorinostat alter BNIP3,Bik, and Bcl-xL protein levels suggest that these three proteinsmight play critical roles in the regulation of apoptosis in meta-static human colon carcinoma cells. The function of BNIP3 inregulating apoptosis in metastatic human colon carcinoma cellswas demonstrated recently (51). To determine whether Bik andBcl-xL function in the apoptosis of colon carcinoma cells, we si-lenced Bcl-xL in LS411N cells and analyzed the tumor cell sen-sitivity to Fas-mediated apoptosis. Analysis of apoptotic cell deathindicated that silencing Bcl-xL significantly increased tumor cellsensitivity to FasL-induced apoptosis (Fig. 4A, 4B). We alsooverexpressed Bik in LS411N cells and observed that restorationof Bik expression significantly increased LS411N cell sensitivityto FasL-induced apoptosis (Fig. 4C, 4D). Therefore, BNIP3, Bik,and Bcl-xL play critical roles in apoptosis in metastatic humancolon carcinoma cells.

Decitabine and vorinostat cooperate to suppress coloncarcinoma development in vivo

To determine whether the above observations can be extended toin vivo colon carcinoma suppression, we made use of the coloncarcinoma CT26 experimental lung metastasis mouse model. Wefirst analyzed the CT26 responses to decitabine and vorinostat andobserved that, like metastatic human colon carcinoma cells, CT26cells responded to decitabine and vorinostat to upregulate Fas andbecame sensitive to FasL-induced apoptosis (Supplemental Fig. 1).Next, CT26 cells were transplanted to syngeneic BALB/c mice.These tumor-bearing mice were then treated with decitabine andvorinostat, either alone or in combination, and examined forlung metastasis. Both decitabine and vorinostat exhibited tumor-suppression effects individually. However, a much greater tumor-suppression effect was observed when decitabine and vorinostatwere used in combination (Fig. 5A). Fas-mediated apoptosis isinitiated by FasL binding to the Fas receptor. To identify thesource of FasL, we first extracted total RNA from lungs derivedfrom tumor-free control mice or tumor-bearing mice and analyzedFasL mRNA levels. Real-time RT-PCR analysis indicated thatboth tumor-free and tumor-bearing lung cells express FasL, andtumor-free lung tissues express higher levels of FasL than do thetumor-bearing lung tissues (Fig. 5B). Next, we sought to deter-mine which types of cells in the lung express FasL. Lungs fromtumor-bearing mice were digested with collagenase to makesingle-cell suspensions and were analyzed for FasL protein levelson the cell surfaces of CD8+ T cells and non-CD8+ cells in thelung. CD8+ T cells accounted for ∼8% of the total lung cells (Fig.5Cb), and ∼10% of lung cells expressed FasL (Fig. 5Ce).Approximately 24.8% of lung tissue-infiltrating CD8+ T cells

expressed FasL (Fig. 5Cc, 5D), whereas ∼12.7% of non-CD8+

cells expressed FasL (Fig. 5Cf, 5D). Taken together, our datasuggest that, although a significant portion of tumor-infiltratingCD8+ T cells is FasL+ cells, both tumor-infiltrating CD8+

T cells and CD82 lung cells are the sources of FasL in the tumormicroenvironment.

FasL plays a critical role in suppression of metastatic coloncarcinoma in vivo

The above observations suggest that FasL is expressed on tumor-infiltrating immune cells, as well as other lung cells. To determinethe role of FasL in tumor suppression in vivo, CT26 cells weretransplanted to wild-type (wt) and Fasgld mice. In the absence ofany treatment, no significant difference in lung tumor burden wasobserved between wt and Fasgld mice (Fig. 6A). However, com-bined decitabine and vorinostat treatment exhibited significantlygreater tumor-suppression efficacy in wt mice than in Fasgld mice(Fig. 6B). Therefore, our results suggest that decitabine and vor-inostat sensitize colon carcinoma cells to FasL-mediated tumorsuppression in vivo.

Low-dose decitabine and vorinostat exhibit no significant livertoxicity in vivo

Toxicity, especially liver toxicity, is the major limitation for the useof DNA methylation inhibitors and HDAC inhibitors in humancancer therapy (29, 31, 52). To determine the toxicity of decitabineand vorinostat at the dose used in this study, we injected the twodrugs i.v. into BALB/c mice, either alone or in combination. Threedays later, blood was collected, and serum was analyzed forcomplete liver enzyme profiles. Aspartate aminotransferase levelwas decreased .2-fold by vorinostat treatment (Table I). Deci-tabine and vorinostat did not significantly alter liver enzymeleakage into the peripheral blood. Taken together, our data suggestthat decitabine and vorinostat exert effective tumor-suppressionactivity at a dose that is not toxic in mice.

A

B gld mice

wt mice No. Tu

mo

r n

od

ule

/mo

use

0

50

100

150

200

250

gld micewt mice

p=0.001

0

20

40

60

80

No. Tu

mo

r n

od

ule

/mo

use

gld micewt mice

p=0.5gld mice

wt mice

wt mice

gld mice

FIGURE 6. Decitabine and vorinostat-mediated tumor suppression is

FasL dependent in vivo. (A) CT26 cells were injected into wt and Fasgld

mice at doses of 7.5 3 104 cells/mouse (upper panel) or 5 3 104 cells/

mouse (lower panel) and examined for lung tumor growth. The number of

lung tumor nodules was enumerated (right panel). Each circle represents

total tumor counts from a single mouse. The horizontal lines in the plot

box represent mean tumor nodule number. (B) CT26 cells were injected

into wt and Fasgld mice at doses of 7.5 3 104 cells/mouse. The tumor-

bearing mice were treated with decitabine and vorinostat, as in Fig. 5A.

Shown are tumor-bearing lungs of wt and Fasgld mice. The number of lung

tumor nodules was enumerated as in (A) (right panel).

6 TARGETING Fas RESISTANCE TO SUPPRESS COLON CANCER

by guest on Decem

ber 19, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

CTL-adoptive immunotherapy in combination with decitabineand vorinostat chemotherapy effectively suppresses coloncarcinoma metastasis

Our above data suggest that decitabine and vorinostat, when used incombination, are effective in overcoming metastatic colon carci-noma cells’ resistance to FasL-induced apoptosis. Our data alsoindicate that FasL plays a critical role in decitabine and vorinostat-mediated tumor suppression in vivo. Because CD8+ T cells ex-press FasL and use FasL as one of their primary effector mech-anisms (9–12), we reasoned that combined chemotherapy withdecitabine and vorinostat and tumor-specific CTL-adoptive im-munotherapy is an effective therapy for the suppression of coloncarcinoma metastasis. To test this hypothesis, CT26 cells weretransplanted into syngeneic mice for 7 d to establish extensivelung metastases. The tumor-bearing mice were then treated withCT26 tumor-specific, perforin-deficient, pfpCTLs or pfpCTLs plusdecitabine and vorinostat. The use of pfpCTLs eliminates theperforin-mediated cytotoxicity to allow better evaluation of theFasL-induced cytotoxicity. The prediction is that if decitabine andvorinostat can overcome apoptosis resistance of the tumor cellsin vivo, then combinational therapy should exhibit greater anti-tumor efficacy than CTL-adoptive immunotherapy alone. Indeed,although combined decitabine and vorinostat treatment (Fig. 5A)and pfpCTLs treatment (Fig. 7) alone showed significant tumor-rejection efficacy, combination chemotherapy plus CTL immu-notherapy exhibited significantly enhanced tumor-rejection effi-cacy against the established colon carcinoma lung metastasescompared with CTL immunotherapy or decitabine and vorinostatchemotherapy alone (Figs. 5A, 7). In summary, our data suggestthat chemotherapy with decitabine and vorinostat in combinationwith CTL adoptive immunotherapy is effective for the interven-tion of colon carcinoma metastasis in vivo.

DiscussionIt is well-established in the literature that decitabine and vorinostatexert direct cytotoxicity to induce tumor cell death, in part throughinducing cell cycle arrest and DNA damage response to activate theintrinsic apoptosis pathway (34, 53, 54). This mechanism mayexplain the decitabine and vorinostat-induced cell death in theabsence of FasL observed in this study (Fig. 1D). Previous studiesalso convincingly demonstrated that vorinostat modulates Fas andother apoptosis-related genes to mediate tumor cell apoptosis (36–38). However, Fas is a death receptor, and it initiates apoptosisonly after engagement by its ligand, FasL. Therefore, an increasein Fas alone and alteration of apoptosis-related genes are notsufficient to initiate Fas-mediated apoptosis in tumor cells in theabsence of FasL. Nevertheless, increased Fas expression levelinduced by decitabine and vorinostat is apparently associated withincreased sensitivity of human colon carcinoma cells to FasL-

induced apoptosis (Fig. 1). More importantly, combined treat-ment with decitabine and vorinostat effectively overcomes meta-static human colon carcinoma cell resistance to Fas-mediatedapoptosis, a hallmark of metastatic human colorectal cancer (27,28).Although a large portion of the tumor-infiltrating CD8+ cells

express FasL, CD82 cells in the lung tissue also express FasL(Fig. 5C). RT-PCR analysis revealed that the FasL mRNA level ishigher in the lungs of tumor-free mice than in those of tumor-bearing mice (Fig. 5B). The majority of cells in lungs of tumor-bearing mice is tumor cells (Fig. 5A). Although tumor-infiltratingCD8+ T cells express FasL, these T cells only consist of a smallportion of totals lung cells. This may explain the high FasL mRNAlevel in the tumor-free mice. The type of FasL+ lung cells andtheir function in Fas-mediated apoptosis require further study.Our studies indicated that the Fas promoter is only sporadically

methylated in metastatic human colon carcinoma cells (Fig. 2).Therefore, decitabine-mediated Fas upregulation is unlikely to

Table I. Liver toxicity profiles in mice after decitabine and vorinostat treatment

Serum Enzyme/Protein Level

Treatment

Control Decitabine Vorinostat Decitabine and Vorinostat

ALP (U/l) 129 120 106 126ALT (U/l) 31 34 18 21AST (U/l) 137 140 54 89Cholesterol (mg/dl) 78 78 76 72Albumin (g/dl) 3.6 3.5 3.5 3.5Total bilirubin (mg/dl) 0.1 0.1 0.1 0.1Total protein (g/dl) 4.8 4.6 4.5 4.4

Serum from three mice of each treatment group was pooled for liver profile analysis.ALP, Alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate phosphatase.

Control

pfpCTLs

pfpCTLs

Decitabine

& Vorinostat

0

50

100

150

200

250

No. T

um

or

nodule

/m

ouse

Control pfpCTLs pfpCTLs

Decitabine

& Vorinostat

p<0.01

p<0.01

p=0.02

A

B

FIGURE 7. Decitabine and vorinostat increase the efficacy of CTL-

adoptive immunotherapy. (A) CT26 cells (2 3 105 cells/mouse) were

injected into syngeneic mice. Seven days later, mice were divided into

three groups. Mice in group 3 were injected with decitabine and vorinostat,

as in Fig. 5A. Then mice in groups 2 and 3 were injected with pfpCTLs

(5 3 105 cells/mouse). Mice in group 3 were treated with decitabine and

vorinostat five times every 2 d. Group 1 mice were used as the control

group. Mice were sacrificed 16 d after tumor transplantation and analyzed

for lung metastasis. (B) The number of lung tumor nodules was enumer-

ated. Each circle represents total counts from a single mouse. The hori-

zontal lines in the plot box represent mean tumor nodule number.

The Journal of Immunology 7

by guest on Decem

ber 19, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

occur through direct inhibition of the Fas promoter DNA meth-ylation. NF-kB and p53 are prominent regulators of Fas expres-sion, and chemotherapeutic agents were shown to upregulate Fasthrough NF-kB– and p53-dependent mechanisms (55, 56) There-fore, it is possible that decitabine and vorinostat upregulate Fasexpression through NF-kB– and p53-dependent mechanisms incolon carcinoma cells, but it remains to be determined.We also demonstrated in this study that BNIP3 and Bik promoter

DNA are methylated in metastatic human colon carcinoma cells(Fig. 3), and decitabine effectively reactivates BNIP3 and Bikexpression in metastatic human colon carcinoma cells. Further-more, vorinostat decreased Bcl-xL expression in metastatic humancolon carcinoma cells. Because silencing Bcl-xL expression oroverexpressing Bik only altered the tumor cell sensitivity to FasL-induced apoptosis to a small degree (Fig. 4), it is likely thatdecitabine and vorinostat cooperate to alter the expression ofmultiple targets, including Fas, BNIP3, Bik, and Bcl-xL, whichadditively contribute to the greater degree of apoptosis inductionin vitro (Fig. 1D) and enhance tumor suppression in vivo (Fig.5A).One of the major obstacles in cancer immunotherapy is immune

suppression. Although tumor-specific FasL+ CTLs are potentiallyeffective anticancer agents (11, 57, 58), the target tumor cellsoften induce immune suppression to suppress CTLs in the tumormicroenvironment (59–61). CTLs suppress target tumor cells pri-marily through two cell contact-dependent cytotoxic mechanisms.The first cytolytic mechanism depends on the polarized secretionof perforin and granzymes. The second effector mechanisminvolves the interaction of FasL on activated CTL surfaces with itsreceptor Fas on the target tumor cells (9, 10, 62, 63). Although itwas shown that regulatory T cells (Tregs) can inhibit clonal ex-pansion of activated T cells in vitro (64), recent studies indicatethat Tregs do not inhibit CD8+ T cell activation and proliferationin vivo but rather selectively inhibit granule exocytosis of CTLs,thereby selectively impairing the perforin effector mechanism ofCTLs without inhibiting CTL activation and clonal expansion (65,66). Therefore, the Fas-mediated cytotoxicity of the tumor-specific CTLs should still be effective and is particularly impor-tant in CTL-mediated antitumor activity under immunosuppres-sive conditions. Furthermore, it was shown that Tregs are highlysensitive to Fas-mediated apoptosis, whereas effector T cells areresistant to Fas-mediated killing (67, 68). Thus, Fas-based cancertherapy may not only induce tumor cell apoptosis, it may alsoinduce Treg apoptosis to eliminate Treg-mediated immune sup-pression. In summary, our data suggest that chemotherapy withdecitabine and vorinostat, in combination with CTL immuno-therapy, is an effective strategy for the suppression of coloncarcinoma metastasis and holds great promise for further devel-opment to treat metastatic colon cancer in human patients.

AcknowledgmentsWe thank Drs. S. Butcher and Lars Damstrup for providing Mega-Fas Li-

gand.

DisclosuresThe authors have no financial conflicts of interest.

References1. Straus, S. E., E. S. Jaffe, J. M. Puck, J. K. Dale, K. B. Elkon, A. Rosen-Wolff,

A. M. Peters, M. C. Sneller, C. W. Hallahan, J. Wang, et al. 2001. The devel-opment of lymphomas in families with autoimmune lymphoproliferative syn-drome with germline Fas mutations and defective lymphocyte apoptosis. Blood98: 194–200.

2. Nikolov, N. P., M. Shimizu, S. Cleland, D. Bailey, J. Aoki, T. Strom,P. L. Schwartzberg, F. Candotti, and R. M. Siegel. 2010. Systemic autoimmunity

and defective Fas ligand secretion in the absence of the Wiskott-Aldrich syn-drome protein. Blood 116: 740–747.

3. Lenardo, M. J., J. B. Oliveira, L. Zheng, and V. K. Rao. 2010. ALPS-ten lessonsfrom an international workshop on a genetic disease of apoptosis. Immunity 32:291–295.

4. Neven, B., A. Magerus-Chatinet, B. Florkin, D. Gobert, O. Lambotte, L. DeSomer, N. Lanzarotti, M. C. Stolzenberg, B. Bader-Meunier, N. Aladjidi, et al.2011. A survey of 90 patients with autoimmune lymphoproliferative syndromerelated to TNFRSF6 mutation. Blood 118: 4798–4807.

5. Dowdell, K. C., J. E. Niemela, S. Price, J. Davis, R. L. Hornung, J. B. Oliveira,J. M. Puck, E. S. Jaffe, S. Pittaluga, J. I. Cohen, et al. 2010. Somatic FASmutations are common in patients with genetically undefined autoimmunelymphoproliferative syndrome. Blood 115: 5164–5169.

6. Takahashi, T., M. Tanaka, C. I. Brannan, N. A. Jenkins, N. G. Copeland, T. Suda,and S. Nagata. 1994. Generalized lymphoproliferative disease in mice, caused bya point mutation in the Fas ligand. Cell 76: 969–976.

7. Siegel, R. M., F. K. Chan, H. J. Chun, and M. J. Lenardo. 2000. The multifacetedrole of Fas signaling in immune cell homeostasis and autoimmunity. Nat.Immunol. 1: 469–474.

8. Owen-Schaub, L. B., K. L. van Golen, L. L. Hill, and J. E. Price. 1998. Fas andFas ligand interactions suppress melanoma lung metastasis. J. Exp. Med. 188:1717–1723.

9. Seki, N., A. D. Brooks, C. R. Carter, T. C. Back, E. M. Parsoneault, M. J. Smyth,R. H. Wiltrout, and T. J. Sayers. 2002. Tumor-specific CTL kill murine renalcancer cells using both perforin and Fas ligand-mediated lysis in vitro, but causetumor regression in vivo in the absence of perforin. J. Immunol. 168: 3484–3492.

10. Caldwell, S. A., M. H. Ryan, E. McDuffie, and S. I. Abrams. 2003. The Fas/Fasligand pathway is important for optimal tumor regression in a mouse model ofCTL adoptive immunotherapy of experimental CMS4 lung metastases. J.Immunol. 171: 2402–2412.

11. Strater, J., U. Hinz, C. Hasel, U. Bhanot, G. Mechtersheimer, T. Lehnert, andP. Moller. 2005. Impaired CD95 expression predisposes for recurrence in cura-tively resected colon carcinoma: clinical evidence for immunoselection andCD95L mediated control of minimal residual disease. Gut 54: 661–665.

12. Schillaci, R., M. Salatino, J. Cassataro, C. J. Proietti, G. H. Giambartolomei,M. A. Rivas, R. P. Carnevale, E. H. Charreau, and P. V. Elizalde. 2006. Immu-nization with murine breast cancer cells treated with antisense oligodeoxynu-cleotides to type I insulin-like growth factor receptor induced an antitumoraleffect mediated by a CD8+ response involving Fas/Fas ligand cytotoxic pathway.J. Immunol. 176: 3426–3437.

13. Park, W. S., R. R. Oh, Y. S. Kim, J. Y. Park, S. H. Lee, M. S. Shin, S. Y. Kim,P. J. Kim, H. K. Lee, N. Y. Yoo, and J. Y. Lee. 2001. Somatic mutations in thedeath domain of the Fas (Apo-1/CD95) gene in gastric cancer. J. Pathol. 193:162–168.

14. Lei, D., E. M. Sturgis, L. E. Wang, Z. Liu, M. E. Zafereo, Q. Wei, and G. Li.2010. FAS and FASLG genetic variants and risk for second primary malignancyin patients with squamous cell carcinoma of the head and neck. Cancer Epi-demiol. Biomarkers Prev. 19: 1484–1491.

15. Yang, M., T. Sun, L. Wang, D. Yu, X. Zhang, X. Miao, J. Liu, D. Zhao, H. Li,W. Tan, and D. Lin. 2008. Functional variants in cell death pathway genes andrisk of pancreatic cancer. Clin. Cancer Res. 14: 3230–3236.

16. Qiu, L. X., J. Shi, H. Yuan, X. Jiang, K. Xue, H. F. Pan, J. Li, and M. H. Zheng.2009. FAS 21,377 G/A polymorphism is associated with cancer susceptibility:evidence from 10,564 cases and 12,075 controls. Hum. Genet. 125: 431–435.

17. Sunter, N. J., K. Scott, R. Hills, D. Grimwade, S. Taylor, L. J. Worrillow,S. E. Fordham, V. J. Forster, G. Jackson, S. Bomken, et al. 2012. A functionalvariant in the core promoter of the CD95 cell death receptor gene predictsprognosis in acute promyelocytic leukemia. Blood 119: 196–205.

18. Sibley, K., S. Rollinson, J. M. Allan, A. G. Smith, G. R. Law, P. L. Roddam,C. F. Skibola, M. T. Smith, and G. J. Morgan. 2003. Functional FAS promoterpolymorphisms are associated with increased risk of acute myeloid leukemia.Cancer Res. 63: 4327–4330.

19. Sung, W. W., Y. C. Wang, Y. W. Cheng, M. C. Lee, K. T. Yeh, L. Wang, J. Wang,C. Y. Chen, and H. Lee. 2011. A polymorphic 2844T/C in FasL promoterpredicts survival and relapse in non-small cell lung cancer. Clin. Cancer Res. 17:5991–5999.

20. Wingender, G., P. Krebs, B. Beutler, and M. Kronenberg. 2010. Antigen-specificcytotoxicity by invariant NKT cells in vivo is CD95/CD178-dependent and iscorrelated with antigenic potency. J. Immunol. 185: 2721–2729.

21. Houghton, J. A., F. G. Harwood, A. A. Gibson, and D. M. Tillman. 1997. The fassignaling pathway is functional in colon carcinoma cells and induces apoptosis.Clin. Cancer Res. 3: 2205–2209.

22. Kagi, D., F. Vignaux, B. Ledermann, K. Burki, V. Depraetere, S. Nagata,H. Hengartner, and P. Golstein. 1994. Fas and perforin pathways as majormechanisms of T cell-mediated cytotoxicity. Science 265: 528–530.

23. Nagata, S. 1997. Apoptosis by death factor. Cell 88: 355–365.24. Huang, D. C., M. Hahne, M. Schroeter, K. Frei, A. Fontana, A. Villunger,

K. Newton, J. Tschopp, and A. Strasser. 1999. Activation of Fas by FasL inducesapoptosis by a mechanism that cannot be blocked by Bcl-2 or Bcl-x(L). Proc.Natl. Acad. Sci. USA 96: 14871–14876.

25. Ogasawara, J., R. Watanabe-Fukunaga, M. Adachi, A. Matsuzawa, T. Kasugai,Y. Kitamura, N. Itoh, T. Suda, and S. Nagata. 1993. Lethal effect of the anti-Fasantibody in mice. Nature 364: 806–809.

26. D’Asaro, M., C. La Mendola, D. Di Liberto, V. Orlando, M. Todaro, M. Spina,G. Guggino, S. Meraviglia, N. Caccamo, A. Messina, et al. 2010. V gamma 9Vdelta 2 T lymphocytes efficiently recognize and kill zoledronate-sensitized,

8 TARGETING Fas RESISTANCE TO SUPPRESS COLON CANCER

by guest on Decem

ber 19, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

imatinib-sensitive, and imatinib-resistant chronic myelogenous leukemia cells. J.Immunol. 184: 3260–3268.

27. Moller, P., K. Koretz, F. Leithauser, S. Bruderlein, C. Henne, A. Quentmeier, andP. H. Krammer. 1994. Expression of APO-1 (CD95), a member of the NGF/TNFreceptor superfamily, in normal and neoplastic colon epithelium. Int. J. Cancer57: 371–377.

28. Krammer, P. H. 1999. CD95(APO-1/Fas)-mediated apoptosis: live and let die.Adv. Immunol. 71: 163–210.

29. Glover, A. B., B. R. Leyland-Jones, H. G. Chun, B. Davies, and D. F. Hoth.1987. Azacitidine: 10 years later. Cancer Treat. Rep. 71: 737–746.

30. Silverman, L. R., E. P. Demakos, B. L. Peterson, A. B. Kornblith, J. C. Holland,R. Odchimar-Reissig, R. M. Stone, D. Nelson, B. L. Powell, C. M. DeCastro,et al. 2002. Randomized controlled trial of azacitidine in patients with themyelodysplastic syndrome: a study of the cancer and leukemia group B. J. Clin.Oncol. 20: 2429–2440.

31. McCarthy, N. 2012. Epigenetics: Worth another look? Nat. Rev. Cancer 12: 2.32. Vansteenkiste, J., E. Van Cutsem, H. Dumez, C. Chen, J. L. Ricker,

S. S. Randolph, and P. Schoffski. 2008. Early phase II trial of oral vorinostat inrelapsed or refractory breast, colorectal, or non-small cell lung cancer. Invest.New Drugs 26: 483–488.

33. Kalac, M., L. Scotto, E. Marchi, J. Amengual, V. E. Seshan, G. Bhagat,N. Ulahannan, V. V. Leshchenko, A. M. Temkin, S. Parekh, et al. 2011. HDACinhibitors and decitabine are highly synergistic and associated with unique gene-expression and epigenetic profiles in models of DLBCL. Blood 118: 5506–5516.

34. Luszczek, W., V. Cheriyath, T. M. Mekhail, and E. C. Borden. 2010. Combi-nations of DNA methyltransferase and histone deacetylase inhibitors induceDNA damage in small cell lung cancer cells: correlation of resistance with IFN-stimulated gene expression. Mol. Cancer Ther. 9: 2309–2321.

35. Ecke, I., F. Petry, A. Rosenberger, S. Tauber, S. Monkemeyer, I. Hess, C. Dullin,S. Kimmina, J. Pirngruber, S. A. Johnsen, et al. 2009. Antitumor effects ofa combined 5-aza-29deoxycytidine and valproic acid treatment on rhabdomyo-sarcoma and medulloblastoma in Ptch mutant mice. Cancer Res. 69: 887–895.

36. Park, M. A., C. Mitchell, G. Zhang, A. Yacoub, J. Allegood, D. Haussinger,R. Reinehr, A. Larner, S. Spiegel, P. B. Fisher, et al. 2010. Vorinostat and sor-afenib increase CD95 activation in gastrointestinal tumor cells through a Ca(2+)-de novo ceramide-PP2A-reactive oxygen species-dependent signaling pathway.Cancer Res. 70: 6313–6324.

37. Park, M. A., G. Zhang, C. Mitchell, M. Rahmani, H. Hamed, M. P. Hagan,A. Yacoub, D. T. Curiel, P. B. Fisher, S. Grant, and P. Dent. 2008. Mitogen-activated protein kinase kinase 1/2 inhibitors and 17-allylamino-17-demethoxy-geldanamycin synergize to kill human gastrointestinal tumor cells in vitro viasuppression of c-FLIP-s levels and activation of CD95. Mol. Cancer Ther. 7:2633–2648.

38. Park, M. A., G. Zhang, A. P. Martin, H. Hamed, C. Mitchell, P. B. Hylemon,M. Graf, M. Rahmani, K. Ryan, X. Liu, et al. 2008. Vorinostat and sorafenibincrease ER stress, autophagy and apoptosis via ceramide-dependent CD95 andPERK activation. Cancer Biol. Ther. 7: 1648–1662.

39. Zhang, G., M. A. Park, C. Mitchell, H. Hamed, M. Rahmani, A. P. Martin,D. T. Curiel, A. Yacoub, M. Graf, R. Lee, et al. 2008. Vorinostat and sorafenibsynergistically kill tumor cells via FLIP suppression and CD95 activation. Clin.Cancer Res. 14: 5385–5399.

40. Kaminskyy, V. O., O. V. Surova, A. Vaculova, and B. Zhivotovsky. 2011.Combined inhibition of DNA methyltransferase and histone deacetylase restorescaspase-8 expression and sensitizes SCLC cells to TRAIL. Carcinogenesis 32:1450–1458.

41. Zimmerman, M., D. Yang, X. Hu, F. Liu, N. Singh, D. Browning, V. Ganapathy,P. Chandler, D. Choubey, S. I. Abrams, and K. Liu. 2010. IFN-g upregulatessurvivin and Ifi202 expression to induce survival and proliferation of tumor-specific T cells. PLoS ONE 5: e14076.

42. Hu, X., D. Yang, M. Zimmerman, F. Liu, J. Yang, S. Kannan, A. Burchert,Z. Szulc, A. Bielawska, K. Ozato, et al. 2011. IRF8 regulates acid ceramidaseexpression to mediate apoptosis and suppresses myelogeneous leukemia. CancerRes. 71: 2882–2891.

43. McGough, J. M., D. Yang, S. Huang, D. Georgi, S. M. Hewitt, C. Rocken,M. Tanzer, M. P. Ebert, and K. Liu. 2008. DNA methylation represses IFN-gamma-induced and signal transducer and activator of transcription 1-medi-ated IFN regulatory factor 8 activation in colon carcinoma cells. Mol. CancerRes. 6: 1841–1851.

44. Liu, F., X. Hu, M. Zimmerman, J. L. Waller, P. Wu, A. Hayes-Jordan, D. Lev,and K. Liu. 2011. TNFa cooperates with IFN-g to repress Bcl-xL expression tosensitize metastatic colon carcinoma cells to TRAIL-mediated apoptosis. PLoSONE 6: e16241.

45. Yang, D., T. J. Stewart, K. K. Smith, D. Georgi, S. I. Abrams, and K. Liu. 2008.Downregulation of IFN-gammaR in association with loss of Fas function islinked to tumor progression. Int. J. Cancer 122: 350–362.

46. Gillenwater, A. M., M. Zhong, and R. Lotan. 2007. Histone deacetylase inhibitorsuberoylanilide hydroxamic acid induces apoptosis through both mitochondrialand Fas (Cd95) signaling in head and neck squamous carcinoma cells. Mol.Cancer Ther. 6: 2967–2975.

47. Petak, I., R. P. Danam, D. M. Tillman, R. Vernes, S. R. Howell, L. Berczi,L. Kopper, T. P. Brent, and J. A. Houghton. 2003. Hypermethylation of the gene

promoter and enhancer region can regulate Fas expression and sensitivity incolon carcinoma. Cell Death Differ. 10: 211–217.

48. Scott, F. L., B. Stec, C. Pop, M. K. Dobaczewska, J. J. Lee, E. Monosov,H. Robinson, G. S. Salvesen, R. Schwarzenbacher, and S. J. Riedl. 2009. TheFas-FADD death domain complex structure unravels signalling by receptorclustering. Nature 457: 1019–1022.

49. Snow, A. L., L. J. Chen, R. R. Nepomuceno, S. M. Krams, C. O. Esquivel, andO. M. Martinez. 2001. Resistance to Fas-mediated apoptosis in EBV-infectedB cell lymphomas is due to defects in the proximal Fas signaling pathway. J.Immunol. 167: 5404–5411.

50. Qin, Y., B. Camoretti-Mercado, L. Blokh, C. G. Long, F. D. Ko, andK. J. Hamann. 2002. Fas resistance of leukemic eosinophils is due to activationof NF-kappa B by Fas ligation. J. Immunol. 169: 3536–3544.

51. Liu, F., Q. Liu, D. Yang, W. B. Bollag, K. Robertson, P. Wu, and K. Liu. 2011.Verticillin A overcomes apoptosis resistance in human colon carcinoma throughDNA methylation-dependent upregulation of BNIP3. Cancer Res. 71: 6807–6816.

52. Ramalingam, S. S., S. Kummar, J. Sarantopoulos, S. Shibata, P. LoRusso,M. Yerk, J. Holleran, Y. Lin, J. H. Beumer, R. D. Harvey, et al. 2010. Phase Istudy of vorinostat in patients with advanced solid tumors and hepatic dys-function: a National Cancer Institute Organ Dysfunction Working Group study.J. Clin. Oncol. 28: 4507–4512.

53. Ruiz-Magana, M. J., J. M. Rodrıguez-Vargas, J. C. Morales, M. A. Saldivia,K. Schulze-Osthoff, and C. Ruiz-Ruiz. 2012. The DNA methyltransferaseinhibitors zebularine and decitabine induce mitochondria-mediated apoptosisand DNA damage in p53 mutant leukemic T cells. Int. J. Cancer 130: 1195–1207.

54. Chen, M. Y., W. S. Liao, Z. Lu, W. G. Bornmann, V. Hennessey, M. N. Washington,G. L. Rosner, Y. Yu, A. A. Ahmed, and R. C. Bast, Jr. 2011. Decitabine andsuberoylanilide hydroxamic acid (SAHA) inhibit growth of ovarian cancer celllines and xenografts while inducing expression of imprinted tumor suppressorgenes, apoptosis, G2/M arrest, and autophagy. Cancer 117: 4424–4438.

55. Muller, M., S. Wilder, D. Bannasch, D. Israeli, K. Lehlbach, M. Li-Weber,S. L. Friedman, P. R. Galle, W. Stremmel, M. Oren, and P. H. Krammer. 1998.p53 activates the CD95 (APO-1/Fas) gene in response to DNA damage by an-ticancer drugs. J. Exp. Med. 188: 2033–2045.

56. Crescenzi, E., F. Pacifico, A. Lavorgna, R. De Palma, E. D’Aiuto, G. Palumbo,S. Formisano, and A. Leonardi. 2011. NF-kB-dependent cytokine secretioncontrols Fas expression on chemotherapy-induced premature senescent tumorcells. Oncogene 30: 2707–2717.

57. Galon, J., A. Costes, F. Sanchez-Cabo, A. Kirilovsky, B. Mlecnik, C. Lagorce-Pages, M. Tosolini, M. Camus, A. Berger, P. Wind, et al. 2006. Type, density, andlocation of immune cells within human colorectal tumors predict clinical out-come. Science 313: 1960–1964.

58. Kilinc, M. O., T. Gu, J. L. Harden, L. P. Virtuoso, and N. K. Egilmez. 2009.Central role of tumor-associated CD8+ T effector/memory cells in restoringsystemic antitumor immunity. J. Immunol. 182: 4217–4225.

59. Petrausch, U., S. M. Jensen, C. Twitty, C. H. Poehlein, D. P. Haley, E. B. Walker,and B. A. Fox. 2009. Disruption of TGF-beta signaling prevents the generationof tumor-sensitized regulatory T cells and facilitates therapeutic antitumor im-munity. J. Immunol. 183: 3682–3689.

60. Lu, T., R. Ramakrishnan, S. Altiok, J. I. Youn, P. Cheng, E. Celis, V. Pisarev,S. Sherman, M. B. Sporn, and D. Gabrilovich. 2011. Tumor-infiltrating myeloidcells induce tumor cell resistance to cytotoxic T cells in mice. J. Clin. Invest.121: 4015–4029.

61. Ostrand-Rosenberg, S., and P. Sinha. 2009. Myeloid-derived suppressor cells:linking inflammation and cancer. J. Immunol. 182: 4499–4506.

62. Medema, J. P., J. de Jong, T. van Hall, C. J. Melief, and R. Offringa. 1999.Immune escape of tumors in vivo by expression of cellular FLICE-inhibitoryprotein. J. Exp. Med. 190: 1033–1038.

63. Liu, K., S. A. Caldwell, K. M. Greeneltch, D. Yang, and S. I. Abrams. 2006. CTLadoptive immunotherapy concurrently mediates tumor regression and tumorescape. J. Immunol. 176: 3374–3382.

64. Shevach, E. M. 2002. CD4+ CD25+ suppressor T cells: more questions thananswers. Nat. Rev. Immunol. 2: 389–400.

65. Chen, M. L., M. J. Pittet, L. Gorelik, R. A. Flavell, R. Weissleder, H. vonBoehmer, and K. Khazaie. 2005. Regulatory T cells suppress tumor-specific CD8T cell cytotoxicity through TGF-beta signals in vivo. Proc. Natl. Acad. Sci. USA102: 419–424.

66. Mempel, T. R., M. J. Pittet, K. Khazaie, W. Weninger, R. Weissleder, H. vonBoehmer, and U. H. von Andrian. 2006. Regulatory T cells reversibly suppresscytotoxic T cell function independent of effector differentiation. Immunity 25:129–141.

67. Fritzsching, B., N. Oberle, N. Eberhardt, S. Quick, J. Haas, B. Wildemann,P. H. Krammer, and E. Suri-Payer. 2005. In contrast to effector T cells, CD4+CD25+FoxP3+ regulatory T cells are highly susceptible to CD95 ligand- butnot to TCR-mediated cell death. J. Immunol. 175: 32–36.

68. Kilinc, M. O., R. B. Rowswell-Turner, T. Gu, L. P. Virtuoso, and N. K. Egilmez.2009. Activated CD8+ T-effector/memory cells eliminate CD4+ CD25+ Foxp3+T-suppressor cells from tumors via FasL mediated apoptosis. J. Immunol. 183:7656–7660.

The Journal of Immunology 9

by guest on Decem

ber 19, 2018http://w

ww

.jimm

unol.org/D

ownloaded from