Efficient genetic modification of murine dendriticcells by electroporation with mRNA

Sonja Van Meirvenne,1 Lieven Straetman,1 Carlo Heirman,1 Melissa Dullaers,1

Catherine De Greef,1 Viggo Van Tendeloo,2 and Kris Thielemans1

1Laboratory of Physiology-Immunology of the Medical School of the Vrije Universiteit Brussel (VUB), Brussels,Belgium; and 2Laboratory of Experimental Hematology, University of Antwerp (UIA), Antwerp UniversityHospital (UZA), Edegem, Belgium.

Recently, human dendritic cells (DCs) pulsed with mRNA encoding a broad range of tumor antigens have proven to be potent

activators of a primary anti– tumor-specific T-cell response in vitro. The aim of this study was to improve the mRNA pulsing of

murine DC. Compared to a standard lipofection protocol and passive pulsing, electroporation was, in our hands, the most efficient

method. The optimal conditions to electroporate murine bone marrow–derived DCs with mRNA were determined using enhanced

green fluorescent protein and a truncated form of the nerve growth factor receptor. We could obtain high transfection efficiencies

around 70–80% with a mean fluorescence intensity of 100–200. A maximal expression level was reached 3 hours after

electroporation. A clear dose–response effect was seen depending on the amount of mRNA used. Importantly, the electroporation

process did not affect the viability nor the allostimulatory capacity or phenotype of the DC. To study the capacity of mRNA-

electroporated DCs to present antigen in the context of MHC classes I and II, we made use of chimeric constructs of ovalbumin. The

dose-dependent response effect and the duration of presentation were also determined. Together, these results demonstrate that

mRNA electroporation is a useful method to generate genetically modified murine DC, which can be used for preclinical studies

testing immunotherapeutic approaches.

Cancer Gene Therapy (2002) 9, 787–797 doi:10.1038/sj.cgt.7700499

Keywords: mRNA electroporation; dendritic cells; antigen presentation; cancer vaccine

Dendritic cells (DCs) are key players in the initiation ofimmune responses and the activation of naive T cells.1

Immature DCs are standing guards widely spread in alltissues. Following antigen capture and appropriate ‘‘danger’’signals, they migrate to the T-cell areas of the draininglymphoid organs. During the migration, DCs lose theirantigen-capturing and -processing activities while acquir-ing exceptional presentation and stimulatory capabilities.2,3

The colocalization of antigen-loaded DCs and T cells spe-cific for the same antigen is likely the very first step in theinduction of an adaptive immune response.

Because of their exceptional potency to activate naive Tcells, DCs have been proposed as the basis of cancervaccines.4–7 Different formulations of tumor antigens havebeen used to load the DC: synthetic peptides,8,9 acid-elutedpeptides,10,11 tumor cell lysates,12,13apoptotic bodies,14 puri-fiedproteins,15–17ornucleic acids. 18–23Theantigensare thenprocessed and presented on the surface of DCs, leading toactivation of T cells.

Because of the limited number of known tumor-associated epitopes and the HLA polymorphism restriction,

a vaccination strategy based upon a genetic approach,whereby the whole tumor antigen (containing the maximalepitopic potential ) is processed and presented, has a lot ofadvantages compared to the more commonly used methodssuch as peptide loading of DCs.24 Recently, human DCspulsed with mRNA encoding a broad range of tumorantigens have proven to be potent activators of a primaryanti– tumor-specific T-cell response in vitro.25–32 Impor-tantly, a successful vaccine should induce antigen-specificCD4+ as well as CD8+ T cells. The antigen-specific CD4+

T cells are needed for the induction and the maintenance ofCD8+ T-cell responses and may directly or indirectlycontribute to tumor cell destruction.33–38 In this context,genetic modification of DCs through chemokines, cytokines,or membrane-bound immunomodulatory molecules couldresult in even more potent immunostimulatory cells.39–42

Several authors have obtained promising results withhuman DCs pulsed with either total tumor RNA or in vitro–transcribed RNA encoding different tumor antigens, e.g.,CEA,PSA,E6, E7, TERT, and yet undefined antigens derivedfrom cancers of different origin.25–28,30–32 Three pulsingtechniques are currently being used, i.e., passive pulsing,lipofection, and electroporation. According to Van Tendelooet al,43 electroporation is the most efficient techniqueto transfect human DCs. As regards genetic modificationof mouse DCs with tumor antigens, similar successfulresults have been reported.25,41,44,45 To our knowledge,

Received May 27, 2002.

Address correspondence and reprint requests to: Kris Thielemans, MD,

PhD, Laboratory of Physiology- Immunology of the Medical School of theVrije Universiteit Brussel (VUB), Laarbeeklaan 103, Brussels B -1090,

Belgium. E-mail: kthielem@fysp.vub.ac.be

Cancer Gene Therapy (2002) 9, 787 – 797D 2002 Nature Publishing Group All rights reserved 0929-1903 /02 $25.00

www.nature.com/cgt

electroporation has not been reported as a pulsing techniquefor mouse DCs.

In this study, we investigated the possibility to geneticallymodify mouse bone marrow–derived DCs by electropora-tion with in vitro–transcribed mRNA. The optimal electro-poration conditions were determined using mRNA codingfor the reporter gene eGFP46 and the truncated NGF-R.47 Toanalyze the antigen presentation by the modified DCs, anmRNA was used for coding of a truncated form of the modelantigen ovalbumin (tOVA),19 modified by the addition ofubiquitin, an intracellular degradation signal,48–50 or by anendolysosomal targeting signal derived from the invariantchain.51–55 We demonstrate that DCs can be efficientlytransfected with mRNA using electroporation and that thechimeric OVA constructs were efficiently processed andtargeted to both the MHC class I and/or class II processingpathways. We also show that mRNA-electroporated DCsare far superior in the antigen presentation assay compared tomRNA-pulsed DCs.

Materials and methods

Mice, cell lines, and reagents

C57BL/6 (H-2b) female mice, 6–8 weeks old, wereobtained from Harlan (Horst, the Netherlands). Animalswere maintained and treated according to the InstitutionalGuidelines.

The tumor cell lines used were EL4 lymphoma (C57BL/6, H-2b) and E.G7-OVA (EL4 cells transfected withchicken OVA cDNA; ATCC, Rockville, MD). EL4 cellswere maintained in DMEM (Invitrogen-Life Technolo-gies, Merelbeke, Belgium) supplemented with 5% heat-inactivated FCS, 2 mM glutamine, 50 �M 2-ME, 100 U/mLpenicillin, and 100 �g/mL streptomycin. E.G7-OVA cellswere grown in the same medium containing 400 �g/mLG418 (Eurobiochem, Bierges, Belgium). The T-cell hybrid-oma RF33.70, which recognizes an H-2-Kb–restrictedOVA peptide (aa 257–264), and the T-cell hybridomaMF2.2D9, which recognizes an H-2-I-Ab–binding epitopelocated within aa 265–280 of the OVA protein sequence,were kindly provided by Dr KL Rock (Boston, MA). Thesecell lines secrete IL-2 upon recognition of the OVA epitopesin the context of the corresponding restriction elements. Theamount of IL-2 was determined using the indicator cell lineCTLL-2. All cell lines used were free of mycoplasma.

Generation of bone marrow–derived DC

Bone marrow–derived DCs were generated following theprotocol described by Lutz et al.56 Briefly, bone marrowcells were isolated from the hind limbs and treated with redblood cell lysis buffer. The cells were plated in a 10-cmbacteriological Petri dish (Falcon-Becton Dickinson, Erem-bodegem, Belgium) at 2�106 cells in 10 mL of completemedium (DMEM supplemented with 5% heat- inactivatedFCS, 2 mM glutamine, 50 �M 2-ME, 100 U/mL penicillin,100 �g/mL streptomycin, and 20 ng/mL rmo GM-CSF).On day 3 of culture, 10 mL of culture medium containing20 ng/mL rmo GM-CSF was added. On day 5, 50% of the

medium was refreshed with culture medium containing20 ng/mL rmo GM-CSF. On day 7, cells were used formRNA electroporation and according to the experimentalset -up DCs were further matured with LPS derived fromEscherichia coli serotype O55:B5 (Sigma-Aldrich, Bornem,Belgium) at a concentration of 0.1 �g/mL.

Plasmids and in vitro transcription of mRNA

The vector pGEM-4Z 50UT-eGFP-30UT-A64 was kindlyprovided by Dr E Gilboa. This plasmid contains a 741-bpeGFP fragment from peGFP-N1 (Clontech, Westburg,Leusden, the Netherlands) flanked by the 50 and 30 UTRsof Xenopus laevis b -globin and 64 A–T bp. Transcription iscontrolled by a bacteriophage T7 promoter. SpeI and NotIsites are present at the 30 end of the A64 stretch to allowlinearization of the plasmid before in vitro transcription. Thetruncated human NGF-R (extracellular and transmembranefragment) was amplified from the plasmid ptNGFR kindlyprovided by Dr C Traversari ( Instituto Scientifico HSRaf-faele, Milan, Italy). The primers used for polymerase chainreaction (PCR) were: 50 -CCCTCATGAGGGCAGGTGC-CACCG-30 and 50 -GGGAGATCTCTAGAGGATCCCCC-TCTTCC-30. During the PCR, a 50 BspHI site spanning thestart codon and a stop codon and a BglII site at the 30 endwere inserted. The eGFP gene was excised from pGEM-50UT-eGFP-30UT-A64 with NcoI and BamHI and replacedwith the BspHI–BglII fragment containing the truncatedNGF-R gene.

The gene fragment coding for the 80 amino-terminalresidues of the murine invariant chain was amplified frompcDNA1moIi80ova, kindly provided by Dr K Frauwirth(University of California, Berkeley, CA). Primers used inPCR were:50 -TTTCCATGGATGACCAGCGCGAC-30 and50 -TTTGGATCCGGAAGCTTCATGCGCAGGTT-30. ThisPCR adds a NcoI site spanning the start codon and a BamHIsite at the 30 end. Ubiquitin was amplified from C57BL/6mouse cDNA using the following primers: 50 -GGGCCAT-GGAAATCTTCGTGAAGACCCTG-30 and 50 -ACACGG-ATCCCCACCGCGCAGACGCAG-30. This PCR adds aNcoI site and a BamHI site at the 50 and 30 ends of the cDNA.

The eGFP gene was excised with NcoI and BamHIand replaced by a NcoI–BamHI fragment containing theamino-terminal part of the mouse invariant chain ( pGEM-moIi80 ). The same strategy was used for the replace-ment of eGFP by ubiquitin ( pGEM-Ubi ). The cDNAcoding for OVA was truncated at its 50 end in order toremove the ER translocation signal. The tOVA protein isnot secreted and mimics a tumor-associated antigen ex-pressed in the cytosol. A BamHI fragment that containstOVA was ligated into BamHI- linearized pGEM-moIi80and pGEM-Ubi vectors, resulting in pGEM-moIi80.tOVAand pGEM-Ubi.tOVA.

For in vitro transcription, the plasmids were linearizedwith either SpeI ( pGEM-eGFP ) or NotI ( pGEM-moIi80.-tOVA and pGEM-Ubi.tOVA ). The linearized plasmidswere used as template for the mMessage mMachine invitro transcription system (Ambion, Austin, TX). Theconcentration and quality of the in vitro–transcribedmRNA were assessed by spectrophotometry and agarosegel electrophoresis.

Cancer Gene Therapy

Electroporation of murine dendritic cellsS Van Meirvenne et al

788

mRNA electroporation, lipofection, and passive pulsingof DC

Immediately before transfection, cells were washed twice inPBS (Invitrogen-Life Technologies) and collected bycentrifugation for 10 minutes at 1500 rpm. The cells wereresuspended in Opti -MEM (Invitrogen-Life Technologies)to a final concentration of 20�106 cells /mL. Subsequently,200 �L of the cell suspension was mixed with 20 �g ofmRNA in a 0.4-cm gap sterile disposable electroporationcuvette and electroporated with an Easyject Plus apparatus(Equibio, Kent, UK). Cells were transfected with a voltagepulse of 300 V in combination with a capacitance of 150 �Fand a pulse time of 6 milliseconds. After electroporation,cells were immediately resuspended in fresh completemedium and further incubated at 378C in a humidifiedatmosphere supplemented with 7% CO2.

Lipofection of DCs was performed following the manu-facturer’s instructions with minor modifications: 20 �g ofmRNA diluted in 250 �L of Opti -MEM was mixed with the

cationic lipid DMRIE-C (Invitrogen-Life Technologies),also diluted in 250 �L of Opti -MEM, at a lipid- to-mRNAratio of 4:1. After incubation for 5–15 minutes at roomtemperature in order to allow mRNA–lipid complexation,lipoplexes were added to 2 mL of DCs suspension in Opti -MEM (5�105 cells /mL) and incubated for 4 hours at 378C.

Alternatively, 2 mL of DCs suspension was passivelypulsed with 20 �g of mRNA in the absence of DMRIE-C for4 hours at 378C. After lipofection or passive pulsing, thecells were further cultured in fresh complete medium at 378Cin a humidified atmosphere supplemented with 7% CO2.

Peptide loading of DC

At day 7, DCs were harvested, washed twice in PBS, andcollected in a 15-mL polystyrene tube in Opti -MEMmedium at a concentration of 4�106 cells /mL. To inducematuration, cells were incubated in a humidified incubatorfor 4 hours in the presence of 0.1 �g/mL LPS. During thelast 2 hours of incubation, a mixture of OVA peptides I-Ab

(TEWTSSNVMEERKIKV) and Kb (SIINFEKL) (Euro-gentec, Seraing, Belgium) was added each at a finalconcentration of 10, 1, or 0.1 �M. After peptide loading,cells were washed twice in medium, resuspended incomplete medium, and replated in a 24-well plate at 106

cells /mL.

Mock

77 % TE152 MFI

eGFP

Figure 1 Transfection efficiency of mRNA-electroporated DCs as

determined by FACS analysis. Twenty- four hours after electro-poration of day 7 DCs with 20 �g of eGFP mRNA, a transfection

efficiency of 77% was obtained with a mean fluorescence intensity

of 152. Similar results were obtained in several (>5) separate

experiments. 00

1010

2020

3030

4040

5050

6060

7070

8080

3h3h 6h6h 12h12h 24h24h 36h36h 48h48h 96h96h

% fl

uore

scen

t pos

itive

cel

l%

fluo

resc

ent p

ositi

ve c

ell ss

eGFPNGFR

00

5050

100100

150150

200200

250250

3h3h 6h6h 12h12h 24h24h 36h36h 48h48h 96h96h

mea

n flu

ores

cent

inte

nsity

mea

n flu

ores

cent

inte

nsity

eGFPNGFR

Figure 2 Transfection efficiency (% positive cells ) and mean

fluorescence intensity of eGFP and truncated NGF-R mRNA-

electroporated DCs followed in time. Day 7 DCs were electroporatedwith 20 �g of mRNA and analyzed by FACS at different time points.

A gradual decrease of both parameters was seen over 96 hours.

Cancer Gene Therapy

Electroporation of murine dendritic cellsS Van Meirvenne et al

789

Flow cytometric analysis

The phenotypic profile of the BM-derived DCs, theexpression of the truncated NGF-R, and the presentationof OVA in the context of MHC class I were analyzed by flowcytometry. The monoclonal antibodies anti -CD11c (N418),anti -B7.1 (16-10A1),57 anti -B7.2 (GL-1), anti -CD40(FGK45),58 anti–I-A/I-E (B21-2), anti–NGF-R (20.4),and anti–Kb–SIINFEKL (25-D1.16)59 were affinity-puri-fied from tissue culture supernatants and conjugated tobiotin. DCs were incubated for 30 minutes with thebiotinylated detection mAb and, after washing, stained withStreptavidin-Cy-Chrome (BD PharMingen, Heidelberg,Germany). All stainings were performed on ice in PBScontaining 1% BSA, 0.02% sodium azide, and 10% 2.4G2mAb supernatant to reduce nonspecific binding through theFcgRII (CD32). Stained cells were fixed in 1% paraformal-dehyde in PBS and analyzed on a FACS Calibur flowcytometer using Cell Quest software (Becton and DickinsonImmunocytometry Systems, San Jose, CA). All stainingswere compared to irrelevant isotype control stainings.

Allogeneic T-cell proliferation

The ability of electroporated DCs to stimulate resting T cellswas assessed by an allo-mixed leukocyte reaction (MLR).

Day 7 C57BL/6 bone marrow–derived DCs were eithernontransfected, mock-electroporated, or electroporated withIi80.tOVA mRNA as described in Materials and Methodsand matured with LPS for 4 hours. The cells were harvestedat day 8, mitomycin C–treated, and plated at graded con-centrations in the presence of 2�105 BALB/c nylonwool purified T cells in 200 �L of DMEM with supplements.After 3 days of incubation at 378C, 5% CO2, the cells werepulsed overnight with 1 �Ci/well [3H]thymidine. Incorpo-ration of [3H]thymidine was measured using a b -scintilla-tion counter (Microbeta-Wallac, Turku, Finland).

Antigen presentation assay

To determine whether the electroporated DCs presentedOVA-derived peptides in the context of either MHC class I orII molecules, OVA-specific T-cell hybridoma cells (5�104

cells /well in a final volume of 200 �L) were cocultured withgraded numbers of DCs harvested at day 8. Cells were platedin round-bottomed 96-well culture plates and incubated at378C in a 5% CO2 humidified incubator. After 20–22 hours,

A

B

0102030

405060708090

elec

trop

orat

ed

Moc

k

1 g

eGF

P

5 g

eGF

P

10

g eG

FP

20

g eG

FP

% fl

uore

scen

t pos

itive

cel

ls

020406080

100120140160180200

non

elec

trop

orat

ed

Moc

k

1 g

eGF

P

5 g

eGF

P

10

g eG

FP

20

g eG

FP

mea

n flu

ores

cenc

e in

tens

ity

non

µ µ µ µ

µ µ µ µ

Figure 3 Effect of the amount of mRNA used for the electroporation

of DCs on the transfection efficiency and on the expression level ofthe reporter gene. Day 7 DCs were electroporated with different

amounts of eGFP mRNA and analyzed by FACS 24 hours later.

A

B

0 20000 40000 60000 80000

Mock electroporated

Ii80.tOVA passive pulsed

Ii80.tOVA lipofected

Ii80.tOVA electroporated

CPM

0 20000 40000 60000 80000

Mock electroporation

Ii80.tOVA passive pulsing

Ii80.tOVA lipofection

Ii80.tOVA electroporation

CPM

Figure 4 Different methods of transfection with OVA-encoding

mRNA compared in an antigen presentation assay. On day 7, DCs

were either mock-electroporated, passively pulsed, lipofected, orelectroporated with Ii80.tOVA mRNA. A: After 24 hours, the

presentation of OVA in the context of MHC class I was determined

by coculture of the transfected DC with RF33.70 TT hybridoma cells.

B: The presentation of OVA in the context of MHC class II wasdetermined using the MF2.2D9 TT hybridoma cells. T cells and DCs

were cocultured at a T:DC ratio of 10:1.

Cancer Gene Therapy

Electroporation of murine dendritic cellsS Van Meirvenne et al

790

the supernatant was harvested and the IL-2 release wasmeasured by the proliferation assay with CTLL-2 cells.

Results

Electroporation of bone marrow–derived DCs within vitro–transcribed mRNA is a highly efficientand nontoxic transfection method

In order to improve the loading of DCs with tumor antigen–encoding mRNA, we investigated the transfection efficacyof electroporation of murine bone marrow–derived DCswith in vitro–transcribed mRNA. We used eGFP -encoding

mRNA as a tool to test different parameters. Flow cytometricanalysis was used to monitor the transfection efficiency andthe level and kinetics of the reporter gene expression.

To address the feasibility of the transfection of murineDCs, we used the same electrical settings as recentlydescribed for the successful electroporation of humanmonocyte–derived DCs.43 Flow cytometric analysis andviability assessment were performed at different time pointsafter electroporation. Figure 1 shows a representative flowcytometric analysis of DCs 24 hours after electroporationwith 20 �g of eGFP -encoding mRNA. Twenty-four hoursafter transfection, 70% (mean of seven experiments ) of theDCs, gated on their forward and sideward scatter character-istics, expressed eGFP with an MFI of 100. Optimaltransfection efficiency was obtained when DCs were usedon day 7 of culture (data not shown). Figure 2 shows thepercentage of positive cells and the fluorescence intensity atdifferent intervals after electroporation. A maximal trans-fection efficiency was detected already 3 hours afterelectroporation and sustained for 24 hours. The expressionlevel reached a maximum 3 hours after electroporation andthen decreased slowly. The efficacy of the electroporationprocess ( i.e., the percentage of eGFP-positive cells ) waswithin the same range when 20, 10, or 5 �g of mRNA wasused. The expression level, however, decreased when loweramounts of mRNA were used (Fig 3). The electroporationprocedure had no effect on the viability of the cells.

To confirm the results obtained with eGFP mRNA, ofwhich only a few molecules are needed to render the cellsfluorescent, we repeated the experiments with mRNAencoding a membrane-bound molecule, namely, a truncatedform of NGF-R. Similar results were obtained (Fig 2).

In summary, electroporation of murine DCs with in vitro–transcribed mRNA is a very efficient and a nontoxic methodfor transient genetic modification of these cells.

Electroporation is much more efficient to introduce mRNAinto DCs than passive pulsing and lipofection

To evaluate the efficacy of various mRNA pulsing methods,presentation of the OVA transgene by C57BL/6 DC, in the

Figure 5 Effect of electroporation upon the maturation status of DCs.

On day 7, DCs were either nontransfected, mock-electroporated, orelectroporated with 20 �g of eGFP mRNA. A FACS staining was

performed 24 hours later to determine the maturation status of the

DCs. An overlay is shown of DCs that are not LPS- treated (grey ) andDCs that are LPS- treated for 24 hours (black).

0 100000 200000 300000 400000

B6 spleen cells

B6 DC non-transfected

B6 DC Mock-electroporated

B6 DC mRNA-electroporated

CPM

Figure 6 Effect of electroporation upon the stimulatory activity of

bone marrow–derived DCs as assessed in a mixed leukocyte

reaction. C57BL/6 spleen cells, LPS-matured DCs, and mock-and Ii80.tOVA mRNA-electroporated and matured DCs were used to

stimulate allogeneic BALB/c T lymphocytes at a T:DC ratio of 40:1.

The proliferation induced by the different DCs populations was

measured by [ 3H]thymidine incorporation after 5 days of culture.Results shown represent the mean of triplicate wells +SEM and are

representative of at least five experiments.

0

20000

40000

60000

80000

100000

120000

Mock Ubi.tOVA Ii80.tOVA

CP

M MHC class IMHC class II

Figure 7 Analysis of the MHC classes I and II presentation of the

OVA transgene by DCs electroporated with either Ubi.tOVAmRNA or

Ii80.tOVA mRNA. Twenty- four hours after electroporation, DCs were

cocultured with the TT hybridomas RF33.70 or MF2.2D9 at a T:DCratio of 10:1. Presentation by Ubi.tOVAmRNA-electroporated DCs is

limited to MHC class I, whereas Ii80.tOVA mRNA-electroporated

DCs exhibit both MHC classes I and II presentation.

Cancer Gene Therapy

Electroporation of murine dendritic cellsS Van Meirvenne et al

791

context of MHC classes I and II, was assessed 24 hours afterpassive pulsing, lipofection, electroporation with Ii80.tOVAmRNA, or mock electroporation. Transfected DCs werecocultured with the TT hybridomas RF33.70 (OVA257–264-Kb–specific) and MF2.2D9 (OVA265–280- I -Ab–specific)at a TT:DCs ratio of 10:1. The amount of IL-2, secretedupon stimulation of these cells, was evaluated in a CTLL-2bioassay. As shown in Figure 4, presentation of OVA, bothin the context of MHC classes I and II, was negligible afterpassive pulsing and very low after lipofection. Thepresentation of OVA by electroporated DCs was 5-foldhigher than by lipofected DCs. The effect of the transfectionmethod on the viability of the DCs was assessed after24 hours by PI staining and FACS analysis. Around 6% ofcell death occurred after passive pulsing, mock electro-poration, and Ii80.tOVA mRNA electroporation, whereas23% cell death occurred after lipofection of DCs (data notshown).

In summary, electroporation of DCs proved a much moreefficient method to introduce mRNA into DCs than passivepulsing and lipofection.

Electroporation does not hamper the maturation potentialnor the allostimulatory capacity

The effect of the electroporation upon the maturation statusof the DCs, their capacity to further mature upon LPSexposure, and, accordingly, their stimulatory capacity toinduce an allo-mixed leukocyte reaction were investigated.First, immature DCs and LPS-treated DCs that were eithernontransfected, mock-electroporated, or electroporated witheGFP mRNA were analyzed by FACS for the expression ofseveral surface markers (Fig 5). The expression of CD11c, aDC-specific marker, was around 80% on day 8 of cultureand comparable for all DCs culture conditions. Importantly,a clear up-regulation of the DCs maturation markers, CD86and CD40, was detected upon a 24-hour LPS stimulation ofeither nontransfected and mock-electroporated DCs as wellas of DCs electroporated with eGFP mRNA. These datademonstrate that electroporation in the absence or presenceof mRNA does not inhibit the maturation capacity of the

A

B

0

10000

20000

30000

40000

50000

60000

24 h 48 h 72 h

CP

M

Mock20 g Ii80.tOVA10 g Ii80. tOVA5 g Ii80.tOVA

0

2000

4000

6000

8000

10000

12000

14000

16000

24 h 48 h 72 h

CP

M

Mock20 g Ii80.tOVA10 g Ii80. tOVA 5 g Ii80. tOVA

µµµ

µµµ

Figure 8 Effect of the amount of mRNA used for the electroporationand the duration of the OVA epitope presentation. A: The

presentation of OVA in the context of MHC class I was determined

through the RF33.70 TT hybridoma cells at the indicated time after

electroporation. B: The presentation of OVA in the context of MHCclass II was determined by means of the MF2.2D9 TT hybridoma

cells. TT hybridoma cells and electroporated DCs were cocultured at

a T:DC ratio of 40:1.

A

B

0

20000

40000

60000

80000

100000

120000

4h 24h 48h

CP

M

MockIi80.tOVABlanco10 M1 M0.1 M

0

20000

40000

60000

80000

100000

120000

4h 24h 48h

CP

M

MockIi80.tOVABlanco10 M1 M0.1 M

µµ

µ

µµ

µ

Figure 9 Comparison of Ii80.tOVA mRNA-electroporated DCs andpeptide-pulsed DCs for their potency to present OVA in the context

of MHC classes I and II. Day 7 DCs were either electroporated with 20

�g of Ii80.tOVA mRNA or loaded with a mixture of OVA265– 280 -Kb

and OVA265 – 280 - I -Ab peptides as described in Materials and

Methods. Cells were further matured with LPS for 4 hours and

screened at different points of time for the presentation of OVA. A:DCs were cocultured with the RF33.70 TT hybridoma cells at a T:DC

ratio of 10:1 or (B) with the MF2.2D9 TT hybridoma cells at a T:DCratio of 40:1.

Cancer Gene Therapy

Electroporation of murine dendritic cellsS Van Meirvenne et al

792

DCs upon LPS stimulation. Moreover, the electroporationitself, in the absence or presence of mRNA, did not inducematuration of the DC. Further, we determined the optimalpoint of time to start the maturation process. A slightly bettermaturation was observed when LPS was added immediatelyafter transfection compared to the addition 24 hours before orafter electroporation (data not shown). We also examinedthe incubation time necessary to induce full maturation ofDCs. FACS analysis for maturation markers showed that a4-hour incubation with 0.1 �g/mL LPS was sufficient ( twoseparate experiments) to induce maximal maturation.

Secondly, an allo-MLR was performed to investigatewhether the electroporation had an effect upon thestimulatory activity of bone marrow DC. C57BL/6 spleencells and 4-hour LPS-treated C57BL/6 DC, which wereeither nontransfected, mock- , or Ii80.tOVA–electroporated,were cocultured with BALB/c T lymphocytes at a T:DCsratio of 40:1 (Fig 6). The proliferation was measured by[3H]thymidine incorporation after 3 days of culture. Nosignificant difference was detected between the differentDCs populations in stimulating the allogeneic T cells ( threeseparate experiments).

In conclusion, the electroporation process neither inhibitsnor induces maturation of DCs, as illustrated by the findingsthat electroporated DCs are equally well matured uponexposure to LPS, and have the same phenotypic character-istics and the same allostimulatory capacity as nontrans-fected DCs treated in a similar way. Genetic modification ofDCs by mRNA electroporation does not hamper theimmunostimulatory capacity of DCs and thus could besuitable for immunotherapeutic purposes.

Electroporated DCs process and present OVA peptides bothin the context of MHC classes I and II

Considering the use of mRNA-electroporated DCs as avaccine modality in cancer immunotherapy, it is of greatimportance that the antigen is processed and efficientlypresented in the context of MHC class I as well as of MHCclass II to allow induction of CD4+ T-cell help for activationand maintenance of tumor-specific CD8+ T cells. In this

context, we constructed two different OVA vectors, namely,pGEM-Ubi.tOVA and pGEM-Ii80.tOVA. As seen in Fig-ure 7, the Ubi.tOVA construct, which is degraded in thecytosol by the proteasome, gave rise to MHC class Ipresentation only, whereas the Ii80.tOVA construct, whichtargets the encoded protein towards the endolysosomalpathway, ensures both MHC classes I and II presentation.Accordingly, we decided to use the Ii80.tOVA construct infurther experiments.

Analogous to the experiments wherein the transfectionefficiency and the expression level (MFI) of eGFP infunction of time were determined, the duration of the OVApresentation in MHC classes I and II after electroporationwas investigated. We also determined whether the amount ofmRNA used for electroporation had an effect on thepresentation level of the DCs. Therefore, DCs were trans-fected with increasing amounts of Ii80.tOVA mRNA and apresentation assay was performed 24, 48, and 72 hours afterelectroporation using either the RF33.70 or the MF2.2D9 TThybridoma cells. As shown in Figure 8, the optimalpresentation of both MHC class I– and MHC class II–restricted OVA epitopes occurred 24 hours after transfectionand then decreased gradually. We observed a clear dose-dependent response effect for the presentation of OVA in thecontext of MHC class I as well as in the context of MHCclass II. In Figure 9, the kinetics of presentation by Ii80.tOVA

Figure 11 Comparison of the expression level of the MHC class I–

SIINFEKL complex between Ii80.tOVA mRNA-electroporated DCs

and SIINFEKL peptide- loaded DCs analyzed by FACS. The mAb

25-D1.16 was used to analyze the expression level of the MHC classI–SIINFEKL complex 24 hours after Ii80.tOVAmRNA electroporation

or SIINFEKL peptide loading of DCs.

0

2

4

6

8

10

12

14

16

18

20

24h 48h 72h

% p

ositi

ve c

ells

Mock20 g Ii80.tOVA10 g Ii80.tOVA 5 g Ii80.tOVA

µµ

µ

Figure 10 Effect of the amount of mRNA used for the electroporationand the duration of the OVA epitope presentation determined by

FACS analysis. The mAb 25-D1.16 was used to analyze the

expression level of the MHC class I–SIINFEKL complex after

Ii80.tOVA mRNA electroporation of DCs.

Cancer Gene Therapy

Electroporation of murine dendritic cellsS Van Meirvenne et al

793

mRNA-transfected DCs was compared to OVA257–264-Kb

and OVA265–280-I -Ab peptide- loaded DCs (three separateexperiments ). Until 24 hours, Ii80.tOVA mRNA-electro-porated DCs were slightly better in stimulating the RF33.70TT hybridoma than peptide-pulsed DC. At 48 hours afterpulsing, no significant difference was observed for MHCclass I presentation between mRNA-electroporated DCs andDCs loaded with 10 �M or 1 �M SIINFEKL peptide. At thispoint of time, the capacity of 0.1 �M SIINFEKL peptide-loaded DCs to stimulate the RF33.70 TT hybridoma wassignificantly decreased. The MHC class II presentation byOVA265–280- I -Ab peptide- loaded DCs was clearly dose-dependent and rapidly declined in time. Only DCs loadedwith 10 �M OVA265–280-I -Ab peptide were more potentstimulators of the MF2.2D9 TT hybridoma compared tomRNA-electroporated DC.

To determine the number of cells presenting the Kb–SIINFEKL complex, electroporated DCs were stained withthe mAb 25-D1.16 and analyzed by FACS. As shown inFigure 10, 24 hours after electroporation of the DCs with20 �g of Ii80.tOVA mRNA, an average of 20% of Kb–SIINFEKL–positive cells was obtained. The expression ofthe Kb–SIINFEKL complex gradually decreased in timeand a clear dose-related response effect was seen. Theseobservations were both in correlation with the presentationassay. Important to notice is that the expression level of theKb–SIINFEKL complex by Ii80.tOVA mRNA-electropo-rated DCs was very low compared to peptide- loaded DCs,as shown in Figure 11. This low FACS staining contrasts tothe highly efficient activation of the OVA-Kb–restrictedhybridoma.

In conclusion, LPS-matured DCs electroporated with theIi80.tOVA construct efficiently present OVA peptides in thecontext of both MHC classes I and II, with an optimalpresentation level 24 hours after transfection.

Discussion

Nowadays, many cancer patients cannot benefit fromrecently developed immunotherapies based upon the use oftumor antigen-loaded DC. Often, the available tumor tissueis too small to prepare tumor lysates or apoptotic bodies andto extract tumor proteins or peptides, which can then be usedfor DCs loading.60 Due to the haplotype restriction, the useof peptides derived from tumor-associated antigens is alsolimited.

Because of the potential broader applicability, we favorthe use of genetically modified DCs.19 Important advantagesrelated to this approach are the independence of the patient’shaplotype,60–63 the potency to present several potentiallyimmunogenic epitopes derived from the endogenouslyexpressed full - length tumor protein, and the possibility tomodify the tumor-encoding cDNA by adding degradation ortargeting signals to promote the processing and presentationof tumor antigen–derived epitopes. Many different methodshave been developed to introduce genetic information intoDCs. Disadvantages encountered after viral transduction ofDCs are toxicity, impaired APC function, and expressionof strong immunogenic viral antigens limiting multiple

immunizations.23,61,63–65 In addition, the use of viral par-ticles as a vehicle for tumor antigen delivery is very time-consuming due to the laborious production, purification, andtitration steps. Boczkowski et al45 have pioneered the use ofRNA to genetically modify DCs. Tumor-derived or invitro–transcribed mRNA as a vehicle for the delivery oftumor antigens has many advantages over the more com-monly used methods.29 The benefits of using mRNA ratherthan DNA comprise its safety and the fact that entry ofmRNA into the cytoplasm is sufficient for transient trans-lation, whereas DNA must enter the nucleus to be tran-scribed and can integrate into the host genome.66 Besides,nonviral DNA transfection methods are inefficient, partic-ularly in nondividing cells.43,67 Another important issueto note is that tumor mRNA can be isolated from a smalltumor burden and can serve as a reliable source of tumorantigen.25 Moreover, the RNA can be amplified in vitroand genetically modified or enriched by subtractivehybridization.25,45,68

The aim of this study was to increase the expression andpresentation level of the tumor antigen after mRNA pulsingof murine bone marrow–derived DCs. Electroporation,using the same electrical settings as recently described byVan Tendeloo et al, proved to be a much more efficientmethod than passive pulsing and lipofection when wecompared the presentation of OVA peptides by the mRNA-modified DCs. It should be mentioned that Gilboa et al havebeen able to obtain impressive cellular immune responses invitro and in vivo against various human and mouse tumorantigens using either lipofected or passively pulsed DCs,although these investigators rarely showed the transfectionefficiency of their method. In our hands, the passive pulsingof DCs was much less efficient compared to electroporationas demonstrated by FACS analysis and antigen presentationassays. This might be due to the amount or concentration ofmRNA or the cell density used in our experiments. We havenot investigated these parameters in detail nor optimizedthe passive pulsing approach. We further optimized theconditions to electroporate murine bone marrow–derivedDCs. Using eGFP as a reporter gene, we obtained trans-fection efficiencies of 70–80% when DCs were electro-porated on day 7 of culture with 20 �g of eGFP mRNA.Similar results were obtained when a truncated form ofNGF-R was used as a reporter gene. Importantly, theelectroporation process did not affect the viability of the DC,their immunophenotype, their capacity to respond to amaturation stimulus, nor their capacity to activate allogeneicT cells in a mixed leukocyte reaction assay. The optimaltransfection efficiency, with the highest mean fluorescentintensity, was detected 3 hours after electroporation and thendecreased gradually. The efficacy of the electroporationprocess was within the same range when 2, 10, or 5 �g ofmRNA was used, whereas the expression level decreasedwhen lower amounts of mRNA were used. In this way, it ispossible to influence the expression level of the transgene inthe mRNA-electroporated DCs. The avidity of the T-cellresponse induced by immunization with DCs electroporatedwith different amounts of mRNA remains to be investigated.Theoretically, it is expected that the avidity of the T-cellresponse will increase when a lower amount of antigen is

Cancer Gene Therapy

Electroporation of murine dendritic cellsS Van Meirvenne et al

794

presented.69–71 This is important because T cells with highavidity are more potent in the recognition and eventualkilling of tumor cells in vivo.

The OVA antigen was chosen as tumor model antigen tostudy the presentation of antigenic peptides by mRNA-electroporated murine bone marrow–derived DCs. Atruncated form of OVA (tOVA) was used to mimic acytosolic expressed tumor antigen. After modification of thetOVA -encoding cDNA with ubiquitin to enhance theproteasome-mediated degradation, electroporated DCs pre-sented OVA peptides only in the context of MHC class I.Modification with the endosomal targeting sequence derivedfrom the invariant chain resulted in presentation of OVApeptides both in the context of MHC classes I and II.Because of the importance of CD4+ T-cell help for theinduction of an effective and long-lasting memory CTLresponse in vivo,33–38 we decided to use Ii80.tOVA mRNAin further experiments. Analogous to the results with eGFP,the optimal presentation of both MHC class I– and MHCclass II– restricted OVA epitopes occurred 4 hours aftertransfection and then decreased gradually. We also couldobserve a clear dose-dependent response effect for thepresentation of both peptides. When, in a TT hybridomaassay, the level of OVA presentation in MHC class I or II byelectroporated DCs was compared to peptide- loaded DCs,no major differences were observed. Until 48 hours,mRNA-electroporated DCs and DCs loaded with 10 �Mand 1 �M SIINFEKL peptide presented the MHC class Iepitope equally well. The expression of the MHC class II–peptide complex was less stable both for mRNA-electro-porated and peptide- loaded DCs compared to the MHCclass I–peptide complex. At all times, DCs loaded with10 �M OVA265–280- I -Ab peptide were remarkably betterstimulators of the MF2.2D9 than mRNA-electroporatedDCs. To determine the number of cells presenting the Kb–SIINFEKL complex and the expression niveau on a singlecell level, a FACS staining with the mAb 25-D1.26 wasperformed. When day 7 DCs were electroporated with 20 �gof Ii80.tOVA mRNA, only 20% of the cells stained positive.Besides this, a clear dose–response effect and a gradualdecrease in time of the Kb–SIINFEKL complex with anoptimum 24 hours after electroporation were observed. Wenoticed that the transfection efficiency, determined in thisway, was much lower compared to the transfection ef-ficiency estimated after eGFP or truncated NGF-R mRNAelectroporation. We assume that the reason for this dis-crepancy might be due to the limited sensitivity of thisassay. An observation that supports this hypothesis is thefact that the Kb-positive cell lines transfected with an OVA -encoding vector, E.G7-OVA, and MO4 cells72 could not bestained with the mAb 25-D1.26. As a control, DCs loadedwith increasing concentrations of SIINFEKL peptide wereincluded. A dose-dependent response effect was observedand a clear positive signal was still detected for DCs loadedwith 0.1 �M SIINFEKL peptide. It should be noted that theconcentration of peptide needed for detection of Kb–SIINFEKL complexes by the 25-D1.26 mAb is muchhigher than the concentration that is required for theactivation of the RF33.70 TT hybridoma cells.73 For thelatter, additional interactions between DCs and T cells such

as costimulatory and adhesion molecules or cytokines cansupport stimulation.

We can conclude that the use of mRNA-electroporatedDCs is a reliable and safe strategy and holds promise forimmunotherapy against cancer. We are now optimizing andevaluating the use of human clinical -grade DCs electro-porated with different tumor-associated antigens to start aclinical trial.

Acknowledgments

We thank Muriel Moser for helpful discussion, Hilde Revetsfor kindly providing the 25-D1.16 hybridoma cells, ElsyVaeremans, Peggy Verbuyst, and Gert De Block for tech-nical assistance, and Yasmina Essam for secretarialassistance. This work was supported by grants to KTfrom the Fund for Scientific Research-Flanders (FWO-Vlaanderen), the Institute for Science and Technology(IWT), the Ministry of Science (IUAP/PAI IV), DeAlgemene Spaar -en Lijfrentekas (ASLK/CGER), and DeBelgische Federatie voor Kankerbestrijding.

References

1. Steinman RM. The dendritic cell system and its role inimmunogenicity. Annu Rev Immunol. 1991;9:271–296.

2. Banchereau J, Steinman RM. Dendritic cells and the control ofimmunity. Nature. 1998;392:245–252.

3. Thery C, Amigorena S. The cell biology of antigen presenta-tion in dendritic cells. Curr Opin Immunol. 2001;13:45–51.

4. Fong L, Engleman EG. Dendritic cells in cancer immunother-apy. Annu Rev Immunol. 2000;18:245–273.

5. Dallal RM, Lotze MT. The dendritic cell and human cancervaccines. Curr Opin Immunol. 2000;12:583–588.

6. Grabbe S, Beissert S, Schwarz T, Granstein RD. Dendritic cellsas initiators of tumor immune responses: a possible strategy fortumor immunotherapy? Immunol Today. 1995;16:117–121.

7. Palucka K, Banchereau J. Dendritic cells: a link between innateand adaptive immunity. J Clin Immunol. 1999;19:12–25.

8. Celluzzi CM, Mayordomo JI, Storkus WJ, Lotze MT, FaloLD Jr. Peptide -pulsed dendritic cells induce antigen-specificCTL-mediated protective tumor immunity. J Exp Med.1996;183:283–287.

9. Porgador A, Snyder D, Gilboa E. Induction of antitumorimmunity using bone marrow–generated dendritic cells. JImmunol. 1996;156:2918–2926.

10. Zitvogel L, Mayordomo JI, Tjandrawan T, et al. Therapy ofmurine tumors with tumor peptide-pulsed dendritic cells:dependence on T cells, B7 costimulation, and T helper cell 1–associated cytokines. J Exp Med. 1996;183:87–97.

11. Nair SK, Boczkowski D, Snyder D, Gilboa E. Antigen-presenting cells pulsed with unfractionated tumor -derivedpeptides are potent tumor vaccines. Eur J Immunol. 1997;27:589–597.

12. Fields RC, Shimizu K, Mule JJ. Murine dendritic cells pulsedwith whole tumor lysates mediate potent antitumor immuneresponses in vitro and in vivo. Proc Natl Acad Sci USA.1998;95:9482–9487.

13. Lambert LA, Gibson GR, Maloney M, Barth RJ Jr. Equipotentgeneration of protective antitumor immunity by variousmethods of dendritic cell loading with whole cell tumorantigens. J Immunother. 2001;24:232–236.

Cancer Gene Therapy

Electroporation of murine dendritic cellsS Van Meirvenne et al

795

14. Albert ML, Sauter B, Bhardwaj N. Dendritic cells acquireantigen from apoptotic cells and induce class I– restrictedCTLs. Nature. 1998;392:86–89.

15. Paglia P, Chiodoni C, Rodolfo M, Colombo MP. Murinedendritic cells loaded in vitro with soluble protein primecytotoxic T lymphocytes against tumor antigen in vivo. J ExpMed. 1996;183:317–322.

16. Inaba K, Metlay JP, Crowley MT, Steinman RM. Dendriticcells pulsed with protein antigens in vitro can prime antigen-specific, MHC-restricted T cells in situ. J Exp Med. 1990;172:631–640.

17. Hsu FJ, Benike C, Fagnoni F, et al. Vaccination of patientswith B-cell lymphoma using autologous antigen -pulseddendritic cells. Nat Med. 1996;2:52–58.

18. Schnell S, Young JW, Houghton AN, Sadelain M. Retrovirallytransduced mouse dendritic cells require CD4+ T cell help toelicit antitumor immunity: implications for the clinical use ofdendritic cells. J Immunol. 2000;164:1243–1250.

19. De Veerman M, Heirman C, Van Meirvenne S, et al. Ret-rovirally transduced bone marrow–derived dendritic cellsrequire CD4+ T cell help to elicit protective and therapeuticantitumor immunity. J Immunol. 1999;162:144–151.

20. Specht JM, Wang G, Do MT, et al. Dendritic cells retrovirallytransduced with a model antigen gene are therapeuticallyeffective against established pulmonary metastases. J Exp Med.1997;186:1213–1221.

21. Song W,Tong Y, Carpenter H, KongHL, Crystal RG. Persistent,antigen-specific, therapeutic antitumor immunity by dendriticcells genetically modified with an adenoviral vector to express amodel tumor antigen. Gene Ther. 2000;7:2080–2086.

22. Kaplan JM, Yu Q, Piraino ST, et al. Induction of antitumorimmunity with dendritic cells transduced with adeno-virus vector–encoding endogenous tumor–associated anti-gens. J Immunol. 1999;163:699–707.

23. Brossart P, Goldrath AW, Butz EA, Martin S, Bevan MJ.Virus -mediated delivery of antigenic epitopes into dendriticcells as a means to induce CTL. J Immunol. 1997;158:3270–3276.

24. Boon T, van der Bruggen P. Human tumor antigens recognizedby T lymphocytes. J Exp Med. 1996;183:725–729.

25. Boczkowski D, Nair SK, Nam JH, Lyerly HK, Gilboa E.Induction of tumor immunity and cytotoxic T lymphocyteresponses using dendritic cells transfected with messengerRNA amplified from tumor cells. Cancer Res. 2000;60:1028–1034.

26. Heiser A, Dahm P, Yancey DR, et al. Human dendritic cellstransfected with RNA encoding prostate - specific antigenstimulate prostate -specific CTL responses in vitro. J Immunol.2000;164:5508–5514.

27. Heiser A, Maurice MA, Yancey DR, et al. Human dendriticcells transfected with renal tumor RNA stimulate polyclonalT-cell responses against antigens expressed by primary andmetastatic tumors. Cancer Res. 2001;61:3388–3393.

28. Heiser A, Maurice MA, Yancey DR, et al. Induction ofpolyclonal prostate cancer–specific CTL using dendritic cellstransfected with amplified tumor RNA. J Immunol. 2001;166:2953–2960.

29. Mitchell DA, Nair SK. RNA-transfected dendritic cells incancer immunotherapy. J Clin Invest. 2000;106:1065–1069.

30. Nair SK, Boczkowski D, Morse M, et al. Induction of primarycarcinoembryonic antigen (CEA)–specific cytotoxic T lym-phocytes in vitro using human dendritic cells transfected withRNA. Nat Biotechnol. 1998;16:364–369.

31. Nair SK, Hull S, Coleman D, et al. Induction of carcinoem-bryonic antigen (CEA)–specific cytotoxic T- lymphocyteresponses in vitro using autologous dendritic cells loaded with

CEA peptide or CEA RNA in patients with metastaticmalignancies expressing CEA. Int J Cancer. 1999;82:121–124.

32. Thornburg C, Boczkowski D, Gilboa E, Nair SK. Induction ofcytotoxic T lymphocytes with dendritic cells transfected withhuman papillomavirus E6 and E7 RNA: implications forcervical cancer immunotherapy. J Immunother. 2000;23:412–418.

33. Toes RE, Ossendorp F, Offringa R, Melief CJ. CD4 T cells andtheir role in antitumor immune responses. J Exp Med.1999;189:753–756.

34. Hung K, Hayashi R, Lafond-Walker A, et al. The central roleof CD4( + ) T cells in the antitumor immune response. J ExpMed. 1998;188:2357–2368.

35. Kalams SA, Walker BD. The critical need for CD4 help inmaintaining effective cytotoxic T lymphocyte responses. J ExpMed. 1998;188:2199–2204.

36. Bennett SR, Carbone FR, Karamalis F, et al. Help forcytotoxic -T-cell responses is mediated by CD40 signalling.Nature. 1998;393:478–480.

37. Ridge JP, Di Rosa F, Matzinger P. A conditioned dendritic cellcan be a temporal bridge between a CD4+ T-helper and a T-killer cell. Nature. 1998;393:474–478.

38. Schoenberger SP, Toes E, van der Voort EI, Offringa R, MeliefCJ. T-cell help for cytotoxic T lymphocytes is mediated byCD40–CD40L interactions. Nature. 1998;393:480–483.

39. Akiyama Y, Watanabe M, Maruyama K, et al. Enhancement ofantitumor immunity against B16 melanoma tumor usinggenetically modified dendritic cells to produce cytokines. GeneTher. 2000;7:2113–2121.

40. Kikuchi T, Moore MA, Crystal RG. Dendritic cells modified toexpress CD40 ligand elicit therapeutic immunity againstpreexisting murine tumors. Blood. 2000;96:91–99.

41. Zhang W, He L, Yuan Z, et al. Enhanced therapeutic efficacyof tumor RNA-pulsed dendritic cells after genetic modificationwith lymphotactin. Hum Gene Ther. 1999;10:1151–1161.

42. Tuting T, Wilson CC, Martin DM, et al. Autologous humanmonocyte–derived dendritic cells genetically modified toexpress melanoma antigens elicit primary cytotoxic T cellresponses in vitro: enhancement by cotransfection of genesencoding the Th1-biasing cytokines IL-12 and IFN-alpha.J Immunol. 1998;160:1139–1147.

43. Van Tendeloo VF, Ponsaerts P, Lardon F, et al. Highly efficientgene delivery by mRNA electroporation in human hemato-poietic cells: superiority to lipofection and passive pulsing ofmRNA and to electroporation of plasmid cDNA for tumorantigen loading of dendritic cells. Blood. 2001;98:49–56.

44. Nair SK, Heiser A, Boczkowski D, et al. Induction of cytotoxicT cell responses and tumor immunity against unrelated tumorsusing telomerase reverse transcriptase RNA transfected den-dritic cells. Nat Med. 2000;6:1011–1017.

45. Boczkowski D, Nair SK, Snyder D, Gilboa E. Dendritic cellspulsed with RNA are potent antigen-presenting cells in vitroand in vivo. J Exp Med. 1996;184:465–472.

46. Cormack BP, Valdivia RH, Falkow S. FACS-optimizedmutants of the green fluorescent protein (GFP). Gene.1996;173:33–38.

47. Mavilio F, Ferrari G, Rossine S, et al. Peripheral bloodlymphocytes as target cells of retroviral vector–mediated genetransfer. Blood. 1994;83:1988–1997.

48. Varshavsky A. The ubiquitin system. Trends Biochem Sci.1997;22:383–387.

49. Varshavsky A. The N-end rule: functions, mysteries, uses.Proc Natl Acad Sci USA. 1996;93:12142–12149.

50. Gonda DK, Bachmair A, Wunning I, et al. Universality andstructure of the N-end rule. J Biol Chem. 1989;264:16700–16712.

Cancer Gene Therapy

Electroporation of murine dendritic cellsS Van Meirvenne et al

796

51. Koch N, van Driel IR, Gleeson PA. Hijacking a chaperone:manipulation of the MHC class II presentation pathway.Immunol Today. 2000;21:546–550.

52. Odorizzi CG, Trowbridge IS, Xue L, et al. Sorting signals inthe MHC class II invariant chain cytoplasmic tail andtransmembrane region determine trafficking to an endocyticprocessing compartment. J Cell Biol. 1994;126:317–330.

53. Pieters J, Bakke O, Dobberstein B. The MHC class II–associated invariant chain contains two endosomal targetingsignals within its cytoplasmic tail. J Cell Sci. 1993;106:831–846.

54. Sanderson S, Frauwirth K, Shastri N. Expression of endogen-ous peptide–major histocompatibility complex class II com-plexes derived from invariant chain–antigen fusion proteins.Proc Natl Acad Sci USA. 1995;92:7217–7221.

55. Sponaas A, Carstens C, Koch N. C- terminal extension of theMHC class II–associated invariant chain by an antigenicsequence triggers activation of naive T cells. Gene Ther. 1999;6:1826–1834.

56. Lutz MB, Kukutsch N, Ogilvie AL, et al. An advanced culturemethod for generating large quantities of highly pure dendriticcells from mouse bone marrow. J Immunol Methods. 1999;223:77–92.

57. Razi -Wolf Z, Galvin F, Gray G, Reiser H. Evidence for anadditional ligand, distinct from B7, for the CTLA-4 receptor.Proc Natl Acad Sci USA. 1993;90:11182–11186.

58. Rolink A, Melchers F, Andersson J. The SCID but not theRAG-2 gene product is required for S mu–S epsilon heavychain class switching. Immunity. 1996;5:319–330.

59. Porgador A, Yewdell JW, Deng Y, Bennink JR, Germain RN.Localization, quantitation, and in situ detection of specificpeptide–MHC class I complexes using a monoclonal antibody.Immunity. 1997;6:715–726.

60. Gilboa E. The makings of a tumor rejection antigen. Immunity.1999;11:263–270.

61. Christ M, Lusky M, Stoeckel F, et al. Gene therapy withrecombinant adenovirus vectors: evaluation of the host immuneresponse. Immunol Lett. 1997;57:19–25.

62. Van den Eynde BJ, van der Bruggen P. T cell defined tumorantigens. Curr Opin Immunol. 1997;9:684–693.

63. Yang Y, Li Q, Ertl HC, Wilson JM. Cellular and humoral

immune responses to viral antigens create barriers to lung-directed gene therapy with recombinant adenoviruses. J Virol.1995;69:2004–2015.

64. Jenne L, Hauser C, Arrighi JF, Saurat JH, Hugin AW. Poxvirusas a vector to transduce human dendritic cells for immunother-apy: abortive infection but reduced APC function. Gene Ther.2000;7:1575–1583.

65. Jonuleit H, Tuting T, Steitz J, et al. Efficient transduction ofmature CD83+ dendritic cells using recombinant adenovirussuppressed T cell stimulatory capacity. Gene Ther. 2000;7:249–254.

66. Luo D, Saltzman WM. Synthetic DNA delivery systems. NatBiotechnol. 2000;18:33–37.

67. Strobel I, Berchtold S, Gotze A, et al. Human dendritic cellstransfected with either RNA or DNA encoding influenza matrixprotein M1 differ in their ability to stimulate cytotoxic Tlymphocytes. Gene Ther. 2000;7:2028–2035.

68. Condon C, Watkins SC, Celluzzi CM, Thompson K, FaloLD Jr. DNA-based immunization by in vivo transfection ofdendritic cells. Nat Med. 1996;2:1122–1128.

69. Alexander -Miller MA, Derby MA, Sarin A, Henkart PA,Berzofsky JA. Supraoptimal peptide–major histocompatibilitycomplex causes a decrease in bc1-2 levels and allows tumornecrosis factor alpha receptor II–mediated apoptosis ofcytotoxic T lymphocytes. J Exp Med. 1998;188:1391–1399.

70. Wherry EJ, Puorro KA, Porgador A, Eisenlohr LC. Theinduction of virus - specific CTL as a function of increasingepitope expression: responses rise steadily until excessivelyhigh levels of epitope are attained. J Immunol. 1999;163:3735–3745.

71. Bullock TN, Colella TA, Engelhard VH. The density ofpeptides displayed by dendritic cells affects immune responsesto human tyrosinase and gp100 in HLA-A2 transgenic mice.J Immunol. 2000;164:2354–2361.

72. Moore MW, Carbone FR, Bevan MJ. Introduction of solubleprotein into the class I pathway of antigen processing andpresentation. Cell. 1988;54:777–785.

73. Kukutsch NA, Rossner S, Austyn JM, Schuler G, Lutz MB.Formation and kinetics of MHC class I–ovalbumin peptidecomplexes on immature and mature murine dendritic cells.J Invest Dermatol. 2000;115:449–453.

Cancer Gene Therapy

Electroporation of murine dendritic cellsS Van Meirvenne et al

797