vaccinationagainsttheforkheadfamilytranscriptionfactorfoxp3 … · associated with reduced risk of...

11
Vaccination against the Forkhead Family Transcription Factor Foxp3 Enhances Tumor Immunity Smita Nair, 1 David Boczkowski, 1 Martin Fassnacht, 3 David Pisetsky, 2 and Eli Gilboa 1 Departments of 1 Surgery, Duke Center for Translational Research and 2 Medicine, Duke University Medical Center, Durham, North Carolina and 3 Department of Internal Medicine I, University of Wuerzburg, Wuerzburg, Germany Abstract Depletion of CD4 + CD25 + regulatory T cells (Treg) by treatment with ACD25 antibody synergizes with vaccination protocols to engender protective immunity in mice. The effectiveness of targeting CD25 to eliminate Treg is limited by the fact that CD25, the low-affinity interleukin-2 receptor, is up-regulated on conventional T cells. At present, foxp3 is the only product known to be exclusively expressed in Treg of mice. However, foxp3 is not expressed on the cell surface and hence cannot be targeted with antibodies. In this study, we tested the hypothesis that vaccination of mice against foxp3, a self- antigen expressed also in the thymus, is capable of stimulating foxp3-specific CTL that will cause the depletion of Treg and enhanced antitumor immunity. Vaccination of mice with foxp3 mRNA-transfected dendritic cells elicited a robust foxp3-specific CTL response and potentiated vaccine-induced protective immunity comparably with that of ACD25 antibody administration. In contrast to ACD25 antibody treatment, repeated foxp3 vaccination did not interfere with vaccine- induced protective immunity. Importantly, foxp3 vaccination led to the preferential depletion of foxp3-expressing Treg in the tumor but not in the periphery, whereas ACD25 antibody treatment led to depletion of Treg in both the tumor and the periphery. Targeting foxp3 by vaccination offers a specific and simpler protocol for the prolonged control of Treg that may be associated with reduced risk of autoimmunity, introducing an approach whereby specific depletion of cells is not limited to targeting products expressed on the cell surface. [Cancer Res 2006;67(1):371–80] Introduction The immune system has established an elaborate network of central and peripheral tolerance mechanisms to discriminate between self and nonself. An important component of this network are the CD4 + CD25 + regulatory T cells (Treg), which mediate self- tolerance and immune homeostasis by acting in a dominant cell extrinsic manner to regulate immune functions (1–3). Two sets of observations also implicate Treg in suppression of tumor immunity. First, Treg accumulation at the tumor sites of cancer patients correlated with disease progression (4, 5). Second, elimination of Treg in mice by treatment with an aCD25 antibody enhances the immune-mediated rejection of tumors (6, 7) and synergizes with vaccination protocols (8, 9). In a recent clinical trial, depletion of Treg in renal cancer patients using an interleukin-2 (IL-2)/diphtheria toxin fusion product (ONTAK) led to enhanced vaccine-induced antitumor immune responses (10). Thus, elimination of Treg could represent an important adjunct to cancer immunotherapy. The low-affinity a-chain IL-2 receptor CD25 is constitutively expressed on Treg and is up-regulated on conventional antigen- activated T cells, including the vaccine-induced antitumor effector T cells. This poses certain limitations on targeting CD25 to deplete Treg: ( a ) Treg rebound following depletion of the CD25-expressing cells within a 3- to 5-week period in mice (11, 12). Indeed, administration of aCD25 antibody subsequent to tumor vaccina- tion was detrimental to vaccine-induced protective immunity (8, 13). Thus, depletion protocols must be confined to a period before vaccination, thus precluding the ability of controlling Tregs over time in the vaccinated patients. (b) Although the thymus is a major source of Treg, Treg can be also generated and/or expanded in the periphery as a result of suboptimal antigenic stimulation (14). Two recent studies have shown that vaccination of tumor- bearing mice led to the de novo conversion (15) or the amplification of preexisting (16) Treg populations, potentially exacerbating tumor-specific immune suppression. Thus, depletion of vaccine-induced Treg would be also beneficial but cannot be accomplished by targeting CD25. (c) Depletion of CD25-expressing cells could potentially interfere with an ongoing T cell–mediated control of subclinical pathogenic infections, with dire consequen- ces that cannot be predicted in advance. (d ) A significant fraction of Treg that, can reach up to 50%, express low to undetectable levels of CD25 (12, 17, 18) and hence will not be subject to elimination by aCD25 antibody therapy. ( e ) In addition, because aCD25 antibody–mediated Treg depletion is nonspecific, it can induce or exacerbate autoimmune pathology (11, 19–21). Other products that are up-regulated on the surface of Treg are GITR, CTLA-4, CD103 or LAG-3, and OX-40, but like CD25, they are not expressed exclusively on Treg (1, 22, 23). Currently, the only gene product known to be exclusively expressed in Treg of mice is Foxp3 (1). Foxp3 is a member of the forkhead/winged-helix family of transcription regulators (24). Mutations in foxp3 leading to multiorgan autoimmune syndromes in mice ( scurfy ) and humans (IPEX syndrome; ref. 24), acquisition of suppressive phenotype by forced expression of foxp3 in conventional T cells (25, 26), and ‘‘knock-in’’ marking studies (17) have shown the pivotal role of foxp3 in determining the immune-suppressive phenotype of Treg. Foxp3 expression is not restricted to CD4 + CD25 + T cells. Foxp3 is also expressed in CD4 + CD25 low/ Treg (12, 17, 18) as well as in subsets of CD8 + T cells exhibiting immune-suppressive properties (27). Thus, targeting foxp3 offers distinct advantages over targeting CD25 to eliminate immune-suppressive cells in vivo . Unlike CD25, foxp3 is a nuclear product and is not expressed on the cell surface. Hence, antibodies or ligand-based reagents, such as ONTAK, cannot be used to eliminate foxp3-expressing cells in vivo . Because CD8 + CTLs can recognize antigenic determinants Requests for reprints: Eli Gilboa, Miller School of Medicine, University of Miami, P.O. Box 019132 (M877), Miami, FL 33101. Phone: 305-243-1767; Fax: 305-243-4409; E-mail: [email protected]. I2007 American Association for Cancer Research. doi:10.1158/0008-5472.CAN-06-2903 www.aacrjournals.org 371 Cancer Res 2007; 67: (1). January 1, 2007 Research Article Research. on July 31, 2020. © 2007 American Association for Cancer cancerres.aacrjournals.org Downloaded from

Upload: others

Post on 06-Jul-2020

2 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: VaccinationagainsttheForkheadFamilyTranscriptionFactorFoxp3 … · associated with reduced risk of autoimmunity, introducing an approach whereby specific depletion of cells is not

Vaccination against the Forkhead Family Transcription Factor Foxp3

Enhances Tumor Immunity

Smita Nair,1David Boczkowski,

1Martin Fassnacht,

3David Pisetsky,

2and Eli Gilboa

1

Departments of 1Surgery, Duke Center for Translational Research and 2Medicine, Duke University Medical Center, Durham,North Carolina and 3Department of Internal Medicine I, University of Wuerzburg, Wuerzburg, Germany

Abstract

Depletion of CD4+CD25+ regulatory T cells (Treg) by treatmentwith ACD25 antibody synergizes with vaccination protocols toengender protective immunity in mice. The effectiveness oftargeting CD25 to eliminate Treg is limited by the fact thatCD25, the low-affinity interleukin-2 receptor, is up-regulatedon conventional T cells. At present, foxp3 is the only productknown to be exclusively expressed in Treg of mice. However,foxp3 is not expressed on the cell surface and hence cannot betargeted with antibodies. In this study, we tested thehypothesis that vaccination of mice against foxp3, a self-antigen expressed also in the thymus, is capable of stimulatingfoxp3-specific CTL that will cause the depletion of Treg andenhanced antitumor immunity. Vaccination of mice withfoxp3 mRNA-transfected dendritic cells elicited a robustfoxp3-specific CTL response and potentiated vaccine-inducedprotective immunity comparably with that of ACD25 antibodyadministration. In contrast to ACD25 antibody treatment,repeated foxp3 vaccination did not interfere with vaccine-induced protective immunity. Importantly, foxp3 vaccinationled to the preferential depletion of foxp3-expressing Treg inthe tumor but not in the periphery, whereas ACD25 antibodytreatment led to depletion of Treg in both the tumor and theperiphery. Targeting foxp3 by vaccination offers a specific andsimpler protocol for the prolonged control of Treg that may beassociated with reduced risk of autoimmunity, introducing anapproach whereby specific depletion of cells is not limited totargeting products expressed on the cell surface. [Cancer Res2006;67(1):371–80]

Introduction

The immune system has established an elaborate network ofcentral and peripheral tolerance mechanisms to discriminatebetween self and nonself. An important component of this networkare the CD4+CD25+ regulatory T cells (Treg), which mediate self-tolerance and immune homeostasis by acting in a dominant cellextrinsic manner to regulate immune functions (1–3). Two sets ofobservations also implicate Treg in suppression of tumorimmunity. First, Treg accumulation at the tumor sites of cancerpatients correlated with disease progression (4, 5). Second,elimination of Treg in mice by treatment with an aCD25 antibodyenhances the immune-mediated rejection of tumors (6, 7) andsynergizes with vaccination protocols (8, 9). In a recent clinicaltrial, depletion of Treg in renal cancer patients using an

interleukin-2 (IL-2)/diphtheria toxin fusion product (ONTAK) ledto enhanced vaccine-induced antitumor immune responses (10).Thus, elimination of Treg could represent an important adjunct tocancer immunotherapy.

The low-affinity a-chain IL-2 receptor CD25 is constitutivelyexpressed on Treg and is up-regulated on conventional antigen-activated T cells, including the vaccine-induced antitumor effectorT cells. This poses certain limitations on targeting CD25 to depleteTreg: (a) Treg rebound following depletion of the CD25-expressingcells within a 3- to 5-week period in mice (11, 12). Indeed,administration of aCD25 antibody subsequent to tumor vaccina-tion was detrimental to vaccine-induced protective immunity(8, 13). Thus, depletion protocols must be confined to a periodbefore vaccination, thus precluding the ability of controlling Tregsover time in the vaccinated patients. (b) Although the thymus is amajor source of Treg, Treg can be also generated and/or expandedin the periphery as a result of suboptimal antigenic stimulation(14). Two recent studies have shown that vaccination of tumor-bearing mice led to the de novo conversion (15) or theamplification of preexisting (16) Treg populations, potentiallyexacerbating tumor-specific immune suppression. Thus, depletionof vaccine-induced Treg would be also beneficial but cannot beaccomplished by targeting CD25. (c) Depletion of CD25-expressingcells could potentially interfere with an ongoing T cell–mediatedcontrol of subclinical pathogenic infections, with dire consequen-ces that cannot be predicted in advance. (d) A significant fractionof Treg that, can reach up to 50%, express low to undetectablelevels of CD25 (12, 17, 18) and hence will not be subject toelimination by aCD25 antibody therapy. (e) In addition, becauseaCD25 antibody–mediated Treg depletion is nonspecific, it caninduce or exacerbate autoimmune pathology (11, 19–21).

Other products that are up-regulated on the surface of Treg areGITR, CTLA-4, CD103 or LAG-3, and OX-40, but like CD25, they arenot expressed exclusively on Treg (1, 22, 23). Currently, the onlygene product known to be exclusively expressed in Treg of mice isFoxp3 (1). Foxp3 is a member of the forkhead/winged-helix familyof transcription regulators (24). Mutations in foxp3 leading tomultiorgan autoimmune syndromes in mice (scurfy) and humans(IPEX syndrome; ref. 24), acquisition of suppressive phenotype byforced expression of foxp3 in conventional T cells (25, 26), and‘‘knock-in’’ marking studies (17) have shown the pivotal role offoxp3 in determining the immune-suppressive phenotype of Treg.Foxp3 expression is not restricted to CD4+CD25+ T cells. Foxp3 isalso expressed in CD4+CD25low/� Treg (12, 17, 18) as well as insubsets of CD8+ T cells exhibiting immune-suppressive properties(27). Thus, targeting foxp3 offers distinct advantages over targetingCD25 to eliminate immune-suppressive cells in vivo .

Unlike CD25, foxp3 is a nuclear product and is not expressed onthe cell surface. Hence, antibodies or ligand-based reagents, suchas ONTAK, cannot be used to eliminate foxp3-expressing cellsin vivo . Because CD8+ CTLs can recognize antigenic determinants

Requests for reprints: Eli Gilboa, Miller School of Medicine, University of Miami,P.O. Box 019132 (M877), Miami, FL 33101. Phone: 305-243-1767; Fax: 305-243-4409;E-mail: [email protected].

I2007 American Association for Cancer Research.doi:10.1158/0008-5472.CAN-06-2903

www.aacrjournals.org 371 Cancer Res 2007; 67: (1). January 1, 2007

Research Article

Research. on July 31, 2020. © 2007 American Association for Cancercancerres.aacrjournals.org Downloaded from

Page 2: VaccinationagainsttheForkheadFamilyTranscriptionFactorFoxp3 … · associated with reduced risk of autoimmunity, introducing an approach whereby specific depletion of cells is not

derived from any cellular compartment expressed on the cellsurface in association with MHC class I molecules, we tested thehypothesis that vaccination of mice against foxp3 is capable ofstimulating a foxp3-specific CTL response leading to theelimination of foxp3-expressing cells and enhanced antitumorimmunity. In this study, we show that despite the fact that foxp3 isa self-antigen expressed in the thymus (25, 26, 28), immunizationof mice against foxp3 elicits a robust foxp3-specific CTL response,enhances vaccine-induced antitumor immunity, and, in contrast toaCD25 antibody treatment, does not interfere with vaccine-induced antitumor immunity. Notably, foxp3 vaccination leads tothe preferential depletion of intratumoral, but not peripheral, Tregthat could translate to reduced risk of autoimmunity.

Materials and Methods

Mice. Four- to 6-week-old C57BL/6 mice (H-2b) were obtained from the

Jackson Laboratory (Bar Harbor, ME). In conducting the research described

in this article, the investigators adhered to the ‘‘Guide for the Care and Use

of Laboratory Animals’’ as proposed by the committee on care ofLaboratory Animal Resources Commission on Life Sciences, National

Research Council. The facilities at the Duke vivarium are fully accredited by

the American Association for Accreditation of Laboratory Animal Care.

Cell lines. The F10.9 clone of the B16 melanoma of C57BL/6 origin is a

highly metastatic, poorly immunogenic, and a low class I expressing cell

line (29). Other cell lines used were EL4 cells (C57BL/6, H-2b, and

thymoma). Cells were maintained in DMEM supplemented with 10% FCS,

25 mmol/L HEPES, 2 mmol/L L-glutamine, and 1 mmol/L sodium pyruvate.

Cloning of murine foxp3, tyrosinase-related protein-2, and actinmRNA and electroporation of dendritic cells. Creation of pSP73-Sph/A64 and cloning of pSP73-Sph/TRP-2/A64 and pGEM4Z/murine actin/A64

were previously described (30). Truncated foxp3 was amplified by reverse

transcription-PCR (RT-PCR) from F10.9 tumor grown in naive C57/B6 miceusing the following primers: 5¶-TATATAAAGCTTGCCACCATGGCTCC-TTCCTTGGCCCTTGGCCCATCC-3¶ and 5¶-ATATATTCTAGACTAGGCGAA-CATGCGAGTAAAC-3¶. This amplifies a fragment from nucleotide 28 of the

coding region, a naturally occurring ATG, to position 1120, mutated fromTAC ! TAG, tyrosine ! stop, thereby disrupting the forkhead that is

encoded by nucleotides 1010 to 1260 and deleting one of two putative

Figure 1. Induction of foxp3-specific CTL in mice vaccinated with foxp3 mRNA–transfected dendritic cells. Dendritic cells (DC ) were transfected with mRNA andinjected into mice. Ten days later, splenocytes were isolated and used in a CTL assay as described in Materials and Methods. A, expression of foxp3. Dendriticcells were transfected with mRNA encoding foxp3 or actin, stained with phycoerythrin-labeled anti-foxp3 antibody, and analyzed by flow cytometry as described inMaterials and Methods. B, CTL recognition of foxp3-expressing dendritic cell targets. Splenocytes from mice immunized with TRP-2 mRNA–transfected dendritic cells,foxp3 mRNA–transfected dendritic cells, or treated with PBS were used in a CTL assay against TRP-2, foxp3, or actin mRNA–transfected dendritic cell targets.C, CTL recognition of foxp3-expressing CD25+ splenocytes. Splenocytes were fractionated into CD25+ and CD25� populations and used as targets in CTL assays withsplenocytes. Splenocytes were first depleted of RBC followed by adherence for 1 h at 37jC. T cells were isolated using the EasySep Mouse T-cell enrichment kitfrom StemCell Technologies (Vancouver, Canada) as per manufacturer’s protocol. T cells were labeled with europium and incubated overnight as described in Materialsand Methods. CD25+ T cells were isolated from the total T cells by positive selection. The cells were first labeled with the aCD25 antibody (PC61) followed by EasySepPE selection kit from StemCell Technologies. The cells that were not CD25+ T cells were used as CD25� T cells for CTL assays. D, susceptibility of CD25+ and CD25�

cell to CTL lysis. Mice were immunized with chicken OVA peptide–pulsed dendritic cells and splenocytes used in a CTL assay against CD25+ cells, CD25� cells, ordendritic cells pulsed with either a class I restricted OVA peptide (SIINFEKL) or a control peptide.

Cancer Research

Cancer Res 2007; 67: (1). January 1, 2007 372 www.aacrjournals.org

Research. on July 31, 2020. © 2007 American Association for Cancercancerres.aacrjournals.org Downloaded from

Page 3: VaccinationagainsttheForkheadFamilyTranscriptionFactorFoxp3 … · associated with reduced risk of autoimmunity, introducing an approach whereby specific depletion of cells is not

nuclear localization signals encoded in nucleotides 1241 to 1270. The

PCR fragment was cloned into pSP73-Sph/64 to create pSP73-Sph/foxp3/

A64. The SpeI-linearized template was used in T7 mMessage mMachine(Ambion, Austin, TX) reactions to produce foxp3 in vitro transcribed RNA.

Dendritic cell electroporation was carried out as previously described (30).

CTL induction in vivo . Bone marrow precursor–derived dendritic cells

were generated and electroporated with RNA as previously described (30).Naive syngeneic mice were immunized s.c. at the base of the ear pinna with

1 to 1.25 � 105 RNA-transfected dendritic cells per ear pinna in 50 AL PBS

for a total of 100 AL per mouse. For experiments testing the combination of

tyrosinase-related protein-2 (TRP-2) and Foxp3 or TRP-2 and actin, micewere immunized with 2 to 2.5 � 105 dendritic cells for each antigen for a

combined 4 to 5 � 105 dendritic cells in 100 AL per mouse. Splenocytes

were harvested 10 days after immunization and depleted of RBC withammonium chloride Tris buffer followed by adherence for 1 h. Non-

adherent splenocytes (107) were cultured with 2 � 105 stimulator cells

(dendritic cells electroporated with RNA) in 5 mL RPMI with 10% FCS,

1 mmol/L sodium pyruvate, 100 IU/mL penicillin, 100 Ag/mL streptomycin,and 5 � 10�5 mol/L h-mercaptoethanol per well in a six-well tissue culture

plate. The responders were stimulated with the same antigen as used for

the immunization. Cells were cultured for 5 days at 37jC and 5% CO2.

Effectors were harvested on day 5 for CTL assay. In vitro cytotoxicity assayswere done as previously described (30).

Tumor immunotherapy. For the F10.9-B16 melanoma model, mice

were implanted with 2.5 � 104 F10.9 cells s.c. (in the flank). Two to 3 days

after tumor implantation, mice were immunized s.c. at the base of each earpinna with 1 to 1.25 � 105 RNA-transfected dendritic cells per ear pinna in

50 AL PBS for a total of 100 AL per mouse. For experiments testing the

combination of TRP-2 and Foxp3 or TRP-2 and actin, mice wereimmunized with 2 to 2.5 � 105 dendritic cells for each antigen for a

combined 4 to 5 � 105 dendritic cells in 100 AL per mouse. Tumor growth

was evaluated every other day starting on day 10. Mice were sacrificed once

the tumor size reached 20 mm. Comparison between two groups was doneusing the log-rank test (Mantel-Haenszel test).

Results

Foxp3 is a self-antigen expressed not only in circulating Treg butalso in the thymus (25, 26, 28). It was, therefore, expected thatfoxp3 will have triggered significant tolerance, and that induction

of foxp3-specific immune responses would be difficult. To test this,we immunized mice against foxp3 using foxp3 mRNA-transfecteddendritic cells, an efficient method of stimulating CD8+ CTLresponses in mice (31). For comparison, mice were immunizedagainst TRP-2, a melanocyte-specific product against which CTLresponses and protective immunity can be routinely stimulated inmice (32). To reduce the possibility that ectopic foxp3 expressioncould adversely affect dendritic cell function, a truncatedfunction-inactivated foxp3 product was used by disrupting theforkhead structure and deleting one of the two putative nuclearlocalization signals (see Materials and Methods). Figure 1A showsthat dendritic cells transfected with foxp3 mRNA but not actinmRNA express foxp3 protein, as measured by flow cytometry.Because the truncated foxp3 product may bind weakly theantibody and/or exhibit reduced stability, the steady-state foxp3levels measured by flow cytometry shown in Fig. 1A are likely torepresent an underestimate of the transfection efficiency andthe generation of T-cell determinants. Figure 1B shows thatimmunization of mice against the truncated foxp3 elicited a CTLresponse, which lysed foxp3 mRNA-transfected dendritic cells, butnot TRP-2 or actin mRNA-transfected dendritic cells. Importantly,the magnitude of foxp3 CTL was comparable with that of TRP-2CTL stimulated in a similar manner. Thus, immunization againstfoxp3 was capable of stimulating a robust foxp3-specific CTLresponse despite the fact that foxp3 is a self-antigen expressed inthe thymus.

Because Treg are capable of directly suppressing the cytolyticfunctions of antigen-activated CD8+ T cells (33, 34), the use ofmRNA-transfected dendritic cells as targets left unanswered thequestion of whether the foxp3-specific CTL shown in Fig. 1B canalso kill their physiologic targets, the Treg. To address this issue,we sorted splenocytes for CD25+ and CD25� cells enriched for Tregand conventional naive T cells, respectively. Using real-time RT-PCR, the CD25+ population that constituted 2% to 3% of totalsplenocytes expressed 45-fold more foxp3 than the CD25�

population (data not shown). Figure 1C shows that theCD25+foxp3+, but not the CD25�foxp3�, cells were killed by thesplenocytes from the foxp3-immunized mice. This experimentshows that Treg are susceptible to lysis by the vaccine-inducedfoxp3 CTL. In vitro killing of the Treg targets (Fig. 1C) was clearlyinferior to that of foxp3 mRNA-transfected dendritic cells (Fig. 1B).This could reflect either inherent differences in the susceptibility ofthe bone marrow–derived dendritic cells and the splenic Treg tolysis by CTL, or direct inhibition of the foxp3 CTL effectorfunctions by the cognate Treg targets. To distinguish betweenthese two possibilities, CD25+ and CD25� T cells, as well asdendritic cells, were pulsed with the class I restricted ovalbumin(OVA) peptide and used as targets in a CTL assays. Splenocytesisolated from mice immunized with OVA peptide–pulsed den-dritic cells were used as effectors. As shown in Fig. 1D , bothfoxp3-expressing CD25+ T cells and ‘‘conventional’’ foxp3-negativeCD25� T cells pulsed with OVA peptide were killed at comparableefficiency, which was significantly lower than that of OVApeptide–pulsed dendritic cells. This result shows that the lowefficiency of foxp3 CTL-mediated lysis of the spleen-derived Tregcompared with the lysis of foxp3 mRNA-transfected dendriticcells is due to reduced susceptibility of splenic T cells to CTL lysiscompared with bone marrow–derived dendritic cells, and not toTreg-mediated inhibition.

Having shown that vaccination against foxp3 induces CTLresponses in vivo , we tested whether co-vaccination of mice against

Figure 2. Inhibition of B16/F10.9 melanoma growth in mice immunizedagainst TRP-2 and foxp3. Mice were implanted s.c. with 2.5 � 104 B16/F10.9tumor cells and, 2 d later, immunized with mRNA-transfected dendritic cellsas shown (10 mice per group). All mice received a constant amount ofmRNA-transfected dendritic cells. Mice were monitored for the appearance ofpalpable tumors. Statistical analysis: actin versus TRP-2 or foxp3, P = 0.002;foxp3 or TRP-2 versus TRP-2 + foxp3, P = 0.01.

Foxp3 Vaccination–Mediated Treg Depletion

www.aacrjournals.org 373 Cancer Res 2007; 67: (1). January 1, 2007

Research. on July 31, 2020. © 2007 American Association for Cancercancerres.aacrjournals.org Downloaded from

Page 4: VaccinationagainsttheForkheadFamilyTranscriptionFactorFoxp3 … · associated with reduced risk of autoimmunity, introducing an approach whereby specific depletion of cells is not

a tumor antigen and foxp3 can potentiate vaccine-inducedantitumor immunity. As a prototype tumor antigen, we targetedTRP-2, which constitutes the dominant tumor antigen in the B16melanoma tumor (32). To test whether foxp3 vaccination enhancesprotective antitumor immunity, mice were implanted s.c. with B16/F10.9 tumor cells, a poorly immunogenic variant of the original B16tumor cell line (29), and immunized 2 days later by a singleinjection of either TRP-2 mRNA–transfected dendritic cells, foxp3mRNA–transfected dendritic cells, or both TRP-2 and foxp-3mRNA–transfected dendritic cells. A control group of mice wasimmunized with actin mRNA–transfected dendritic cells. Tumorgrowth was monitored by determining the time to appearance ofpalpable tumors. As shown in Fig. 2, a single immunization withTRP-2 mRNA–transfected dendritic cells resulted in a significantdelay in tumor growth, but eventually all mice succumbed totumor. Likewise, vaccination against foxp3 also resulted in amodest delay in tumor growth similar to that of vaccinationagainst TRP-2. However, when TRP-2 vaccination was combined

with foxp3 vaccination a significant inhibition of tumor growth wasseen, a proportion of mice remaining tumor-free long term (3 of 10mice in Fig. 2). These observations support the view that removal ofTreg in the tumor-bearing mice by vaccination against foxp3enhances both a naturally occurring, albeit weak, antitumorresponse and synergizes with antitumor vaccination. The magni-tude of the antitumor response resulting in the apparent cure of aproportion of the tumor-bearing animals following a singlevaccination cycle is indicative of the benefits of combining asuboptimal antitumor vaccination protocol with vaccinationagainst foxp3.

Current methods used to deplete Treg in vivo target CD25 bysystemic administration of aCD25 antibody. The experimentshown in Fig. 3A compares the antitumor effects of foxp3vaccination (left) and aCD25 antibody treatment (right), showingthat both treatments elicit an essentially identical response asalso seen in Fig. 2 (i.e., a significant retardation, but no long-term survival when foxp3 vaccination or antibody is given to

Figure 3. Induction of protective immunityin mice vaccinated against foxp3 or treatedwith aCD25 antibody. A, enhancementof tumor immunity in mice vaccinatedagainst foxp3 or treated withaCD25antibody (Ab). Mice (10 mice per group)were implanted with B16/F10.9 tumor cellsand immunized against TRP-2 and foxp3as described in Fig. 2 (left) or treated withaCD25 antibody clone PC61 (Serotec,Raleigh, NC) by injection of 400 Ag PC61antibody (Serotec) i.p. on the same day asimmunization with the dendritic cells (right ).Statistical analysis: TRP-2 versus TRP-2 +foxp3, P = 0.045; foxp3 versus TRP-2 +foxp3, P = 0.015; TRP-2 versus TRP-2 +aCD25 antibody, P = 0.03; aCD25antibody versus TRP-2 + aCD25 antibody,P = 0.04; TRP-2 + foxp3 versus TRP-2 +aCD25 antibody, P = 0.8. B, induction ofTRP-2 CTL in mice vaccinated againstfoxp3 or treated with aCD25 antibody. Micewere immunized against TRP-2 and eitherco-vaccinated against foxp3 or treated withaCD25 antibody. Splenocytes wereisolated 7 d later, and the presence ofTRP-2–specific CTL was determinedusing TRP-2–expressing B16/F10.9 andTRP-2–negative EL4 cells as targets, asdescribed in Fig. 1 and Materials andMethods.

Cancer Research

Cancer Res 2007; 67: (1). January 1, 2007 374 www.aacrjournals.org

Research. on July 31, 2020. © 2007 American Association for Cancercancerres.aacrjournals.org Downloaded from

Page 5: VaccinationagainsttheForkheadFamilyTranscriptionFactorFoxp3 … · associated with reduced risk of autoimmunity, introducing an approach whereby specific depletion of cells is not

actin-vaccinated mice, and potentiation of TRP-2 vaccination-induced tumor inhibition with a fraction of mice remainingtumor-free). Although conditions for either antibody depletion orfoxp3 vaccination have not been optimized, these results suggestthat the antitumor effect of foxp3 vaccination is comparablewith that of aCD25 antibody treatment. To determine whetherenhanced protective immunity correlates with CTL responses,induction of TRP-2–specific CTL was determined in ex vivostimulated splenocytes using a standard cytotoxicity assay. Asshown in Fig. 3B , despite the fact that TRP-2 is a self-antigen,even a single vaccination elicited, albeit a modest, CTL response,which killed TRP-2–expressing B16/F10.9 tumor cells, but notTRP-2–negative EL4 tumor cells (left). As expected, CD25antibody–mediated depletion of Treg led to a significant enhance-ment of the vaccine-induced TRP-2–specific CTL response(middle). Consistent with our hypothesis and the data shownin Figs. 2 and 3A , co-immunization against foxp3 also resulted inan enhanced CTL response, which was comparable with thatseen in the antibody treated mice (right). To determine the roleof CD4+ and CD8+ T-cell subsets in protective immunity whenvaccinated mice are co-vaccinated against foxp3 or given aCD25antibody, CD4+ or CD8+ T cells were depleted with corre-sponding antibodies 1 week after immunization. Figure 4 showsthat both CD4+ and CD8+ T cells were necessary to conferprotective immunity in the TRP-2–vaccinated mice depleted of

Treg by either co-vaccination against foxp3 or aCD25 antibodyadministration.

Following depletion, Tregs eventually rebound and reach normallevels within 3 to 5 weeks (11, 12). Moreover, the tumor (15) or thevaccination protocol itself (16) could contribute to the de novogeneration and/or further expansion of the Treg pool. Thus, from atherapeutic standpoint, regulating Treg levels over time byperiodically repeating the depletion protocol will be highlydesirable. We hypothesized that because foxp3 expression, unlikeCD25 expression, is restricted to Treg, repeated foxp3 vaccinationshould not interfere with a vaccine-induced antitumor T-cellresponse. To test this hypothesis, B16/F10.9 tumor-bearing micewere vaccinated against TRP-2 and foxp3 2 days after tumorimplantation and monitored for tumor growth. Another group ofmice received a second foxp3 vaccination 1 week after the firstvaccination (9 days after tumor implantation). For comparison,instead of foxp3 vaccination, mice were treated with aCD25antibodies once or twice using the same schedule. Consistent withthe results shown in Fig. 3A , a single treatment with aCD25antibody (Fig. 5A) or vaccination against foxp3 (Fig. 5B) combinedwith TRP-2 vaccination was superior to either antibody treatmentor foxp3 vaccination given in the absence of TRP-2 vaccination.(As shown in Fig. 3A , aCD25 antibody administration or foxp3vaccination alone had a modest antitumor effect, which wascomparable with that of TRP-2 vaccination). As was previously

Figure 4. The contribution of CD4+ andCD8+ T cells to protective immunity. B16/F10.9 tumor-bearing mice were immunizedagainst TRP-2 and either co-vaccinatedagainst foxp3 or treated with an aCD25antibody (10 mice per group), as describedin Figs. 2 and 3A and Materials andMethods. At day 7 after vaccination andaCD25 antibody administration, mice weretreated with a depleting dose (500 Ag) ofaCD4 antibody clone 2.43 (A ) or aCD8antibody clone GK1.5 (B ) or isotypecontrols, as shown. Statistical analysis:TRP-2 + aCD25 antibody,aCD4 antibodyversus isotype antibody, P = 0.002;TRP-2 + foxp3,aCD4 antibody versusisotype antibody, P = 0.007; TRP-2 +aCD25 antibody,aCD8 antibody versusisotype antibody, P = 0.002; TRP-2 +foxp3,aCD8 antibody versus isotypeantibody, P = 0.013.

Foxp3 Vaccination–Mediated Treg Depletion

www.aacrjournals.org 375 Cancer Res 2007; 67: (1). January 1, 2007

Research. on July 31, 2020. © 2007 American Association for Cancercancerres.aacrjournals.org Downloaded from

Page 6: VaccinationagainsttheForkheadFamilyTranscriptionFactorFoxp3 … · associated with reduced risk of autoimmunity, introducing an approach whereby specific depletion of cells is not

shown (8, 13), a second administration of aCD25 antibody 7 daysafter TRP-2 vaccination led to accelerated tumor growth (Fig. 5A),most likely due to depletion of the vaccine-activated TRP-2 T cells.In contrast to aCD25 treatment, a second foxp3 vaccination hadno adverse effect on the vaccine-induced protective antitumorresponse, although it did not result in better protection (Fig. 5B).In a second experiment, a slight enhancement of protectiveimmunity was seen following a second foxp3 vaccination, but it didnot reach statistical significance (data not shown). A likely reasonwhy a second foxp3 vaccination did not result in furtherenhancement of antitumor immunity is that it takes 3 to 5 weeksfor Treg to rebound (11, 12), coupled with the aggressive growth ofthe B16/F10.9 tumors.

The fate of foxp3-expressing Treg in the periphery and thetumors of mice vaccinated against either TRP-2, TRP-2 + aCD25antibody treatment, or TRP-2 + foxp3 was determined by multi-parameter flow cytometry. As shown in Fig. 6A , administration ofaCD25 antibody resulted in a 2-fold reduction of foxp3-expressingcells in the periphery, measured in the spleen or the vaccine-draining lymph node. As expected, the majority of foxp3-expressingcells in the lymph node or spleen were CD4+ T cells. Notably, andconsistent with recent findings (12, 17, 18), not all CD4+foxp3+

T cells expressed CD25. In fact, the 50% reduction in foxp3-expressing cells following aCD25 antibody administration wasdue to the fact that a significant fraction of the CD4+foxp3+ cellsexpressed low to undetectable levels of CD25 and not becauseaCD25 antibody depletion was inefficient, underscoring one of thelimitations of using aCD25 antibody to control Treg activity in vivo .Surprisingly, as shown in Fig. 6B , foxp3 vaccination did not affectthe percentage of foxp3-expressing cells in the draining lymph

node or the spleen. We have consistently seen that in the spleen,but not the lymph nodes, of the foxp3-vaccinated mice, a numericalredistribution of CD25+foxp3+ to CD25/�foxp3+ has taken place.The mechanism and significance of this phenomenon is currentlybeing investigated.

The effect of foxp3 vaccination or aCD25 antibody treatment ontumor-infiltrating foxp3-expressing cells is shown in Fig. 6C . Bothfoxp3 vaccination or aCD25 antibody treatment led to a significantreduction in the numbers of intratumoral foxp3-expressing cells(85% and 80%, respectively). Both treatments also resulted in amodest 2-fold decrease in tumor infiltrating CD4+ T cells, mostlyreflecting the depletion of the foxp3+CD4+ T cells. Interestingly,foxp3 vaccination, and to a lesser extent aCD25 antibodytreatment, was also accompanied by a decrease in ( foxp3 negative)CD8+ T cells in the tumor. Importantly, however, both methods ofTreg depletion resulted in f2-fold increase in the ratio ofconventional T cells/Treg in the tumors of the treated mice(Fig. 6D). In summary, flow cytometric analysis has shown thataCD25 antibody treatment and foxp3 vaccination result in theintratumoral depletion of Treg, but whereas aCD25 antibodytreatment also affected the peripheral pool of Treg, foxp3vaccination did not. Thus, foxp3 vaccination, in stark contrast toaCD25 antibody treatment, seems to cause the preferentialdepletion of Treg in the tumor, but not in the periphery.

Foxp3 vaccination or CD25 depletion protocols used in thisstudy were not associated with morbidity, mortality, or signs ofvitiligo, and no anti-DNA antibody were detected in the foxp3-immunized mice (data not shown). It remains to be seen if thedifferential effect of foxp3 vaccination on tumor-infiltrating Tregwill translate to reduced autoimmune manifestations.

Figure 5. Effect on tumor immunity whenfoxp3 vaccination or aCD25 antibody aregiven subsequent to TRP-2 vaccination.A, tumor bearing mice were immunizedagainst TRP-2 and treated withaCD25 antibody as described in Fig. 4and Materials and Methods. A secondgroup of mice were retreated with aCD25antibody 7 d after vaccination, and acontrol group of mice was vaccinatedagainst actin and treated with aCD25antibody. B, same as (A ), except thataCD25 antibody treatment was replacedwith foxp3 vaccination. Statistical analysis:actin + aCD25 antibody versus TRP-2 +aCD25 antibody, P = 0.006; TRP-2 +aCD25 antibody versus TRP-2 + aCD25antibody (2�), P = 0.01; actin + foxp3versus TRP-2 + foxp3, P = 0.003.

Figure 6. Flow cytometry analysis of foxp3-expressing cells in mice vaccinated against foxp3 or treated with aCD25 antibody. A and B, mice were vaccinated againstTRP2 and either depleted of CD25+ cells with antibodies or co-vaccinated against foxp3 as described in Materials and Methods and Fig. 4. Seven to 10 d later,spleens and auricular lymph nodes were harvested, and cells were stained with antibodies against foxp3, CD25, and CD4 and analyzed by flow cytometry. Cellswere labeled for surface markers for 20 min on ice with aCD25 antibody/APC and aCD4 antibody/FITC before intracellular staining for Foxp3. Staining for Foxp3 wasdone with the phycoerythrin anti-mouse Foxp3 antibody staining kit and according to the manufacturer’s protocol. All antibodies and isotype controls were obtainedfrom eBioscience (San Diego, CA). Foxp3high cells (left) as gated populations were analyzed for CD4 and CD25 expression (right ). C, mice were implanted s.c. with B16melanoma cells immediately before foxp3 vaccination or aCD25 antibody treatment, as described in Fig. 2 legend. Tumors were harvested as soon as they werepalpable and pooled (Trp-2 group, 4 of 5 tumors; Trp-2 + foxp3 group, 4 of 5 tumors; Trp-2 + aCD25 antibody group, 3 of 5 tumors). Single tumor cell suspensions weregenerated, and cells were stained with APC-conjugated aCD25 antibody and FITC-conjugated aCD4 antibody and phycoerythin/Cy5–conjugated aCD8, fixed,permeabilized, and stained with phycoerythrin-conjugated anti-foxp3 antibody. D, ratio of conventional T cells (Foxp3+CD4+ and CD8+ cells) to Treg cells (Foxp3+ cells)in the pooled tumors of the three treatment groups was calculated from the flow cytometry data shown in (C ).

Cancer Research

Cancer Res 2007; 67: (1). January 1, 2007 376 www.aacrjournals.org

Research. on July 31, 2020. © 2007 American Association for Cancercancerres.aacrjournals.org Downloaded from

Page 7: VaccinationagainsttheForkheadFamilyTranscriptionFactorFoxp3 … · associated with reduced risk of autoimmunity, introducing an approach whereby specific depletion of cells is not

Foxp3 Vaccination–Mediated Treg Depletion

www.aacrjournals.org 377 Cancer Res 2007; 67: (1). January 1, 2007

Research. on July 31, 2020. © 2007 American Association for Cancercancerres.aacrjournals.org Downloaded from

Page 8: VaccinationagainsttheForkheadFamilyTranscriptionFactorFoxp3 … · associated with reduced risk of autoimmunity, introducing an approach whereby specific depletion of cells is not

Discussion

The underlying hypothesis of this study was that immunizationagainst foxp3 would potentiate vaccine-induced antitumorimmunity. We have shown that (a) vaccination of mice withfoxp3 mRNA–transfected dendritic cells elicits a robust foxp3-specific CTL response, which is capable of killing foxp3-expressing Tregs in vitro (Fig. 1); (b) co-vaccination of miceagainst a tumor antigen (TRP-2) and foxp3 potentiates thevaccine-induced protective immunity (Figs. 2 and 3A), whichcorrelates with enhanced induction of TRP-2–specific CTLresponses (Fig. 3B); and (c) the effect of foxp3 vaccination ontumor immunity is comparable with that of aCD25 antibodyadministration (Fig. 3), the current gold standard in Tregdepletion–mediated enhancement of tumor immunity (6–9).Nevertheless, despite having established the concept thatinduction of foxp3-specific CTL responses can enhance tumorimmunity, neither the foxp3 vaccination protocol nor the aCD25antibody treatment have been optimized to permit a conclusionabout their relative potency. Considering the aggressive natureand low immunogenicity of the B16/F10.9 melanoma tumor cellline, the antitumor effects seen in the tumor-bearing miceimmunized once without boosting against an endogenous ‘‘self ’’tumor antigen and foxp3 (Figs. 2 and 3A) is not unremarkable.In vivo elimination of T cells by vaccinating against T cell–specificproducts is not without precedence. Cohen et al. have shown thatvaccination with irradiated T cells corresponding to a T-cell clonespecific to a myelin basic protein (MBP) epitope can protect micefrom MBP-induced experimental autoimmune encephalomyelitis.Notably, the T-cell response was not directed against idiotypicdeterminants of the MBP T-cell clone but rather to determinantsexpressed on activated T cells; yet, no adverse effects were seen inthe immunized mice (35, 36).

Foxp3 is a self-antigen that is also expressed in the thymus, bothin thymocytes destined to become Treg (25, 26) and in thymicstromal cells (28). It was, therefore, surprising that we did notdetect significant tolerance against this thymic antigen. As shownin Fig. 1B , immunizing mice against foxp3 elicits a CTL response ofa magnitude comparable with that of a CTL response inducedagainst TRP-2. In preliminary experiments, foxp3-specific CTLresponses could be also stimulated in vitro from the PBMC ofhuman volunteers, suggesting that the absence of tolerance tofoxp3 may also extend to human settings (data not shown).Although providing evidence for lack of operational toleranceagainst foxp3 in the CD8+ T-cell compartment, which can beexploited for vaccination purposes, these experiments do notexclude the existence of effective tolerance in vivo especially in theCD4+ T-cell and humoral compartments.

Because CD25 is also up-regulated on conventional T cells uponactivation, targeting CD25 to deplete Treg must be restricted to aperiod before vaccination, precluding the long-term control of Treg,which in mice rebound within 3 to 5 weeks of depletion (11, 12).Indeed, administration of aCD25 antibody following vaccinationabrogated protective immunity conceivably by depleting thevaccine-induced tumor-specific effector T cells (refs. 8, 13 and thisstudy; Fig. 5A). A main advantage of targeting foxp3 is that foxp3expression is highly restricted to Treg and, at least in mice, isnot expressed in antigen-activated T cells (1). Consistent with theTreg-restricted expression of foxp3 and in stark contrast to aCD25antibody administration, a second foxp3 vaccination subsequentto tumor vaccination did not adversely affect the vaccine-induced

protective immune response (Fig. 5B), suggesting that targetingfoxp3 will offer a strategy to control Treg over time. The recentdemonstration that the vaccination protocol itself can generatede novo and/or amplify a preexisting Treg pool (15, 16) underscoresthe potential benefits of regulating Treg function after vaccinationwhen the tumor-specific conventional T cells express CD25.

Whether foxp3 expression is restricted to Treg in humans isless clear. Several studies have shown that foxp3 expressionis up-regulated in polyclonally activated ‘‘conventional’’CD4+CD25� T cells (37–39), whereas one study that used apurified naive CD4+CD45RA+CD25� T cell subset failed toobserve up-regulation of foxp3 expression upon antigenicstimulation (40). Yet, another study has reported a 10-folddifference in foxp3 expression between the polyclonally activatedconventional T cells and Treg (41). A recent study has shownthat a proportion (<25%) of polyclonally activated conventional Tcells up-regulated foxp3 transiently and at significantly lowerlevels compared with Treg (42). Thus, at present, the weight ofevidence favors the view that the detection of foxp3 in ex vivocultured human T cells represents either the up-regulation offoxp3 in precommitted Treg, the expansion of a small number ofpreexisting foxp3+ Treg, or the transient and low level up-regulation of foxp3 in a fraction of T cell receptor–stimulatedconventional T cells (which could possibly reflect suboptimalin vitro stimulation conditions favoring the de novo generationof Treg, as shown in murine systems; ref. 14). However, thepossibility that in humans, in stark contrast to rodents, foxp3represents an activation marker necessary but not sufficient toconfer a suppressive phenotype, cannot be excluded.

A second potential advantage of targeting foxp3 for Tregdepletion is that foxp3 expression defines a broader set of immunesuppressive cells, including a subset of CD8+ T cells exhibitingsuppressive potential (27). In particular, a significant fraction ofthe foxp3-expressing CD4+ Treg-expressing low to undetectablelevels of CD25 yet retains suppressive functions (12, 17, 18, 26). Inhumans, CD25 expression on CD4+ T cells as determined by flowcytometry does not provide a clear demarcation betweenconventional T cells and Treg (43). This will require the calibrationof the depletion intensity, a procedure that may be difficult toachieve in clinical settings and could vary from patient to patient(10). The existence of CD4+CD25�Foxp3+ Treg explains whyeffective depletion of >90% of CD4+CD25+ Treg using aCD25antibody correlated with only a modest 2- to 3-fold depletion offoxp3+ cells (Fig. 6A).

Flow cytometry analysis has shown that foxp3 vaccination, incontrast to aCD25 antibody depletion, leads to the preferentialdepletion of Treg from the tumor (Fig. 6). Both methods targetingTreg were accompanied by a significant reduction in the numberof tumor infiltrating foxp3-expressing cells (Fig. 6C). However,whereas aCD25 antibody treatment led to a 2- to 3-fold depletionof Treg in the periphery as measured in the lymph nodes andspleen (Fig. 6A) and the bone marrow (data not shown),vaccination against foxp3 had no effect on the number of Tregin the periphery (Fig. 6B). aCD25 antibody treatment can induce,and more often exacerbate, autoimmunity (11, 19–21) conceivablybecause the across-the-board depletion of Treg consistent withthe 2- to 3-fold depletion of peripheral Treg shown in Fig. 6A . Thedifferential depletion of Treg limited to the tumor by foxp3vaccination raises the possibility that foxp3 vaccination will beassociated with reduced risk of autoimmunity. Future studies willtest this hypothesis. What could be the reasons for this differential

Cancer Research

Cancer Res 2007; 67: (1). January 1, 2007 378 www.aacrjournals.org

Research. on July 31, 2020. © 2007 American Association for Cancercancerres.aacrjournals.org Downloaded from

Page 9: VaccinationagainsttheForkheadFamilyTranscriptionFactorFoxp3 … · associated with reduced risk of autoimmunity, introducing an approach whereby specific depletion of cells is not

effect? One possibility is that the activated foxp3-specific T cellswill preferentially home to the tumor sites (enriched for foxp3expressing Treg) according to the same principles that theactivated T cells home to the inflamed sites and tumors. This isreminiscent of studies showing that activated CTL targeted toantigens expressed in the tumor and peripheral tissue can inhibittumor growth with minimal or no adverse effect on the normaltissues (44–46). A second reason could be that Treg in the tumortissue are in an activated state having recently encountered tumorantigens, and, as reported, activated Treg up-regulate Foxp3expression up to 5- to 10-fold (25, 47–50), thus becoming moresusceptible for CTL recognition and elimination than theirquiescent peripheral counterparts. Future studies will need toexplore the mechanism underlying these observations.

Because mutations in foxp3 are the primary defect in multi-organ autoimmune syndromes in mice (scurfy) and humans (IPEXsyndrome; ref. 24), effective, even if transient, depletion of foxp3-expressing cells, which will encompass a broader spectrum ofimmune suppressive cells than CD25-targeted protocols, couldhave significant adverse effects. The foxp3 vaccination or CD25depletion protocols used in this study were not associated withsigns of morbidity or mortality. No signs of vitiligo were seen in thetreated C57BL/6 mice, nor were anti-DNA antibody generated inthe Treg-depleted mice. Nevertheless, a concern associated withactive immunotherapy is that it could lead to uncontrolledpersistence of (anti-foxp3) immunity and long-term depressedlevels of Treg. This, however, is not likely because a sustainedimmune response is dependent on continued access to antigen ina proinflammatory context, which, in this instance, is provided,and hence can be controlled, by vaccination. A specific concern ofvaccinating against foxp3 is that it is expressed in the thymus, andthat vaccination against foxp3 could adversely affect thymicfunctions. Thus, elimination or reduction, even if transient, of

foxp3-expressing cells by active immunotherapy could haveundesirable adverse effects and will require careful examinationusing optimized and extended Treg depletion protocols.

As an adjunct to immunotherapy, targeting foxp3 by vaccina-tion is a highly suitable protocol for depleting Treg in clinicalsettings. Targeting cell surface products using antibodies orligand-based reagents requires the development, optimization,and validation of a separate procedure and often relies on reagentsthat are not easily accessible for clinical use. On the other hand,foxp3 vaccination does not involve a separate procedure becausepatients would be co-vaccinated against foxp3 and tumorantigen(s) using the same protocol. In this study, we used mRNA-transfected dendritic cells to stimulate immunity against tumorand foxp3 (31). There is no reason to think that the foxp3vaccination approach will not apply to other vaccination strategiesas well. Thus, pending further optimizations and careful testing foradverse effect, targeting foxp3 by vaccination could rapidly progressto clinical testing.

In summary, in this study, we described a strategy to depletefoxp3-expressing cells in vivo by stimulating a foxp3-specific CTLresponse. Unlike antibody- or ligand-based methods, stimulatingCTL responses against products exhibiting differential expressionpattern represents a broadly applicable method to selectivelydeplete cells in vivo , which is not limited to targeting productsexpressed on the cell surface.

Acknowledgments

Received 8/4/2006; revised 10/31/2006; accepted 11/3/2006.Grant support: NIH/National Cancer Institute grants R01 CA102500 and R01

CA098637.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.

References1. Fontenot JD, Rudensky AY. A well adapted regulatorycontrivance: regulatory T cell development and theforkhead family transcription factor Foxp3. Nat Immu-nol 2005;6:331–7.

2. Sakaguchi S. Naturally arising Foxp3-expressingCD25+CD4+ regulatory T cells in immunologicaltolerance to self and non-self. Nat Immunol 2005;6:345–52.

3. von Boehmer H. Mechanisms of suppression bysuppressor T cells. Nat Immunol 2005;6:338–44.

4. Curiel TJ, Coukos G, Zou L, et al. Specific recruitmentof regulatory T cells in ovarian carcinoma fostersimmune privilege and predicts reduced survival. NatMed 2004;10:942–9.

5. Sato E, Olson SH, Ahn J, et al. Intraepithelial CD8+

tumor-infiltrating lymphocytes and a high CD8+/regu-latory T cell ratio are associated with favorableprognosis in ovarian cancer. Proc Natl Acad Sci U S A2005;102:18538–43.

6. Shimizu J, Yamazaki S, Sakaguchi S. Induction oftumor immunity by removing CD25+CD4+ T cells: acommon basis between tumor immunity and autoim-munity. J Immunol 1999;163:5211–8.

7. Onizuka S, Tawara I, Shimizu J, et al. Tumor rejectionby in vivo administration of anti-CD25 (interleukin-2receptor alpha) monoclonal antibody. Cancer Res 1999;59:3128–33.

8. Sutmuller RP, van Duivenvoorde LM, van Elsas A, et al.Synergism of cytotoxic T lymphocyte-associated antigen4 blockade and depletion of CD25(+) regulatory T cellsin antitumor therapy reveals alternative pathways for

suppression of autoreactive cytotoxic T lymphocyteresponses. J Exp Med 2001;194:823–32.

9. Comes A, Rosso O, Orengo AM, et al. CD25+ regulatoryT cell depletion augments immunotherapy of micro-metastases by an IL-21-secreting cellular vaccine. JImmunol 2006;176:1750–8.

10. Dannull J, Su Z, Rizzieri D, et al. Enhancement ofvaccine-mediated antitumor immunity in cancerpatients after depletion of regulatory T cells. J ClinInvest 2005;115:3623–33.

11. Jones E, Dahm-Vicker M, Simon AK, et al. Depletionof CD25+ regulatory cells results in suppression ofmelanoma growth and induction of autoreactivity inmice. Cancer Immun 2002;2:1–8.

12. Zelenay S, Lopes-Carvalho T, Caramalho I, et al.Foxp3+ CD25� CD4 T cells constitute a reservoir ofcommitted regulatory cells that regain CD25 expressionupon homeostatic expansion. Proc Natl Acad Sci U S A2005;102:4091–6.

13. Ko K, Yamazaki S, Nakamura K, et al. Treatment ofadvanced tumors with agonistic anti-GITR mAb and itseffects on tumor-infiltrating Foxp3+CD25+CD4+ regula-tory T cells. J Exp Med 2005;202:885–91.

14. Jaeckel E, Kretschmer K, Apostolou I, et al.Instruction of Treg commitment in peripheral T cellsis suited to reverse autoimmunity. Semin Immunol2006;18:89–92.

15. Valzasina B, Piconese S, Guiducci C, et al.Tumor-induced expansion of regulatory T cells byconversion of CD4+CD25� lymphocytes is thymusand proliferation independent. Cancer Res 2006;66:4488–95.

16. Zhou G, Drake CG, Levitsky HI. Amplification of

tumor-specific regulatory T cells following therapeuticcancer vaccines. Blood 2006;107:628–36.

17. Fontenot JD, Rasmussen JP, Williams LM, et al.Regulatory T cell lineage specification by the fork-head transcription factor foxp3. Immunity 2005;22:329–41.

18. Ono M, Shimizu J, Miyachi Y, et al. Control ofautoimmune myocarditis and multiorgan inflamma-tion by glucocorticoid-induced TNF receptor fam-ily-related protein(high), Foxp3-expressing CD25+

and CD25� regulatory T cells. J Immunol 2006;176:4748–56.

19. Ait-Oufella H, Salomon BL, Potteaux S, et al. Naturalregulatory T cells control the development of athero-sclerosis in mice. Nat Med 2006;12:178–80.

20. Morgan ME, Sutmuller RP, Witteveen HJ, et al. CD25+

cell depletion hastens the onset of severe disease incollagen-induced arthritis. Arthritis Rheum 2003;48:1452–60.

21. Wei WZ, Jacob JB, Zielinski JF, et al. Concurrentinduction of antitumor immunity and autoimmunethyroiditis in CD4+ CD25+ regulatory T cell-depletedmice. Cancer Res 2005;65:8471–8.

22. Fontenot JD, Rudensky AY. Molecular aspects ofregulatory T cell development. Semin Immunol 2004;16:73–80.

23. Sakaguchi S. Naturally arising CD4+ regulatoryt cells for immunologic self-tolerance and negativecontrol of immune responses. Annu Rev Immunol2004;22:531–62.

24. Ramsdell F, Ziegler SF. Transcription factors inautoimmunity. Curr Opin Immunol 2003;15:718–24.

25. Fontenot JD, Gavin MA, Rudensky AY. Foxp3

Foxp3 Vaccination–Mediated Treg Depletion

www.aacrjournals.org 379 Cancer Res 2007; 67: (1). January 1, 2007

Research. on July 31, 2020. © 2007 American Association for Cancercancerres.aacrjournals.org Downloaded from

Page 10: VaccinationagainsttheForkheadFamilyTranscriptionFactorFoxp3 … · associated with reduced risk of autoimmunity, introducing an approach whereby specific depletion of cells is not

programs the development and function ofCD4+CD25+ regulatory T cells. Nat Immunol 2003;4:330–6.

26. Hori S, Nomura T, Sakaguchi S. Control of regulatoryT cell development by the transcription factor Foxp3.Science 2003;299:1057–61.

27. Vlad G, Cortesini R, Suciu-Foca N. License to heal:bidirectional interaction of antigen-specific regulatoryT cells and tolerogenic APC. J Immunol 2005;174:5907–14.

28. Chang X, Gao JX, Jiang Q, et al. The Scurfy mutationof FoxP3 in the thymus stroma leads to defectivethymopoiesis. J Exp Med 2005;202:1141–51.

29. Porgador A, Tzehoval E, Vadai E, et al. Combinedvaccination with major histocompatibility class I andinterleukin 2 gene-transduced melanoma cells syner-gizes the cure of postsurgical established lung metasta-ses. Cancer Res 1995;55:4941–9.

30. Nair S, Boczkowski D, Moeller B, et al. Synergybetween tumor immunotherapy and antiangiogenictherapy. Blood 2003;102:964–71.

31. Gilboa E, Vieweg J. Cancer immunotherapy withmRNA-transfected dendritic cells. Immunol Rev 2004;199:251–63.

32. Schreurs MW, Eggert AA, de Boer AJ, et al.Dendritic cells break tolerance and induce protectiveimmunity against a melanocyte differentiation antigenin an autologous melanoma model. Cancer Res 2000;60:6995–7001.

33. Chen ML, Pittet MJ, Gorelik L, et al. Regulatory T cellssuppress tumor-specific CD8 T cell cytotoxicity throughTGF-beta signals in vivo . Proc Natl Acad Sci U S A 2005;102:419–24.

34. Mempel TR, Pittet MJ, Khazaie K, et al. RegulatoryT cells reversibly suppress cytotoxic T cell functionindependent of effector differentiation. Immunity 2006;25:129–41.

35. Ben-Nun A, Wekerle H, Cohen IR. Vaccination againstautoimmune encephalomyelitis with T-lymphocyte linecells reactive against myelin basic protein. Nature 1981;292:60–4.

36. Cohen IR, Quintana FJ, Mimran A. Tregs in T cellvaccination: exploring the regulation of regulation. J ClinInvest 2004;114:1227–32.

37. Morgan ME, van Bilsen JH, Bakker AM, et al.Expression of FOXP3 mRNA is not confined toCD4+CD25+ T regulatory cells in humans. Hum Immu-nol 2005;66:13–20.

38. Verhasselt V, Vosters O, Beuneu C, et al. Induction ofFOXP3-expressing regulatory CD4pos T cells by humanmature autologous dendritic cells. Eur J Immunol 2004;34:762–72.

39. Walker MR, Kasprowicz DJ, Gersuk VH, et al.Induction of FoxP3 and acquisition of T regulatoryactivity by stimulated human CD4+CD25� T cells. J ClinInvest 20003;112:1437–43.

40. Yagi H, Nomura T, Nakamura K, et al. Crucial role ofFOXP3 in the development and function of humanCD25+CD4+ regulatory T cells. Int Immunol 2004;16:1643–56.

41. Allan SE, Passerini L, Bacchetta R, et al. The role of 2FOXP3 isoforms in the generation of human CD4 Tregs.J Clin Invest 2005;115:3276–84.

42. Gavin MA, Torgerson TR, Houston E, et al. Y.Single-cell analysis of normal and FOXP3-mutanthuman T cells: FOXP3 expression without regulatory

T cell development. Proc Natl Acad Sci U S A 2006;103:6659–64.

43. Baecher-Allan C, Viglietta V, Hafler DA. HumanCD4+CD25+ regulatory T cells. Semin Immunol 2004;16:89–98.

44. Bronte V, Apolloni E, Ronca R, et al. Genetic vac-cination with ‘‘self ’’ tyrosinase-related protein 2 causesmelanoma eradication but not vitiligo. Cancer Res 2000;60:253–8.

45. Melero I, Singhal MC, McGowan P, et al. Immuno-logical ignorance of an E7-encoded cytolytic T-lympho-cyte epitope in transgenic mice expressing the E7 andE6 oncogenes of human papillomavirus type 16. J Virol1997;71:3998–4004.

46. Morgan DJ, Kreuwel HT, Fleck S, et al. Activation oflow avidity CTL specific for a self epitope results intumor rejection but not autoimmunity. J Immunol 1998;160:643–51.

47. Nishikawa H, Kato T, Tawara I, et al. Definition oftarget antigens for naturally occurring CD4(+) CD25(+)regulatory T cells. J Exp Med 2005;201:681–6.

48. Prakken BJ, Samodal R, Le TD, et al. Epitope-specificimmunotherapy induces immune deviation of proin-flammatory T cells in rheumatoid arthritis. Proc NatlAcad Sci U S A 2004;101:4228–33.

49. Yu P, Lee Y, Liu W, Krausz T, et al. Intratumordepletion of CD4+ cells unmasks tumor immunogenicityleading to the rejection of late-stage tumors. J Exp Med2005;201:779–91.

50. Zou L, Barnett B, Safah H, et al. Bone marrow is areservoir for CD4+CD25+ regulatory T cells that trafficthrough CXCL12/CXCR4 signals. Cancer Res 2004;4:8451–5.

Cancer Research

Cancer Res 2007; 67: (1). January 1, 2007 380 www.aacrjournals.org

Research. on July 31, 2020. © 2007 American Association for Cancercancerres.aacrjournals.org Downloaded from

Page 11: VaccinationagainsttheForkheadFamilyTranscriptionFactorFoxp3 … · associated with reduced risk of autoimmunity, introducing an approach whereby specific depletion of cells is not

2007;67:371-380. Cancer Res   Smita Nair, David Boczkowski, Martin Fassnacht, et al.   Factor Foxp3 Enhances Tumor ImmunityVaccination against the Forkhead Family Transcription

  Updated version

  http://cancerres.aacrjournals.org/content/67/1/371

Access the most recent version of this article at:

   

   

  Cited articles

  http://cancerres.aacrjournals.org/content/67/1/371.full#ref-list-1

This article cites 48 articles, 25 of which you can access for free at:

  Citing articles

  http://cancerres.aacrjournals.org/content/67/1/371.full#related-urls

This article has been cited by 7 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] at

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

  Permissions

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

.http://cancerres.aacrjournals.org/content/67/1/371To request permission to re-use all or part of this article, use this link

Research. on July 31, 2020. © 2007 American Association for Cancercancerres.aacrjournals.org Downloaded from