Bioscience Reports, Vol. 19, No. 6, 1999

Intradermal DNA Immunization: Antisera Specific for theMembrane Lectin MR60/ERGIC-53

Violaine Carriere,1 Ludovic Landemarre,2 Valerie Altemayer,3 Genevieve Motta,1Michel Monsigny,1 and Annie-Claude Roche1,4

Received April 15, 1999; accepted August 16, 1999

The trafficking of intracellular membrane proteins in Golgi apparatus, endoplasmic reticu-lum or intermediate compartment has not yet been fully elucidated. The human MR60/ERGIC-53 and the rat p58 proteins are one such protein; and to study them in cell-freeand in situ systems, high quality monospecific antisera are required. Highly specific antiserahave been obtained after immunization of mice with plasmids containing a gene encodingeither the full length or a truncated protein. The best results were obtained after intradermalinjections of a plasmid encoding a truncated protein comprising both the luminal carbo-hydrate recognition domain and the stem down to a cysteine residue close to the C-terminalend, but neither the transmembrane nor the cytosolic domains. Such antisera have a veryhigh titer and are very efficient tools to visualize the MR60 protein in situ or to selectivelyprecipitate the MR60 proteins from a whole cell lysate.

KEY WORDS: Confocal microscopy; ERGIC-53; Human lectin; Mannose-specific lectin;Mouse antibody; MR60.

ABBREVIATIONS: BSA, bovine serum albumin; DMEM, Dulbecco's modified Eagle'smedium; DTT, dithiothreitol; EDTA, ethylenediamine tetraacetic acid; ER, endoplasmicreticulum; ERGIC, endoplasmic reticulum-Golgi-intermediate compartment; FBS, fetalbovine serum; FTC, fluoresceinylthiocarbamyl; PCR, polymerase chain reaction; PEI,polyethyleneimine; PFA, paraformaldehyde; PBS, phosphate buffered saline; PMSF,phenylmethylsulphonyl fluoride; SDS, sodium dodecyl sulfate; TRTC, tetramethylrhodaminylthiocarbamyl.

INTRODUCTION

The need for highly specific antibodies directed against cellular proteins remainsstrong; especially to characterize subcellular compartments or to study intracellulartraffic. The preparation of a membrane protein in a pure form and in amount suf-ficient to immunize animals is difficult. As an alternative, DNA immunization couldprovide an elegant method to produce monospecific antibodies.

1Glycobiologie, Centre de Biophysique Moleculaire, CNRS, Rue Charles Sadron, F-45071 Orleans cedex02, France. E-mail: roche@cnrs-orleans.fr

2Agro-Bio, La Chavannerie, rue Denis Papin, F-45240, La Ferte Saint Aubin.3Present address, Laboratoire de Parasitologie, UFR des Sciences Universite d'Orleans, BP 6759, F-45067Orleans cedex 2.

4To whom correspondence should be addressed.

5590144-8463/99/1200-0559$16.00/0 © 1999 Plenum Publishing Corporation

560 Carriere, Landemarre, Altemayer, Motta, Monsigny, and Roche

A few years ago, we purified and characterized a new intracellular membranelectin (sugar binding protein, MR60) from human cells. This protein is a mannose-specific lectin (Pimpaneau et al., 1991) belonging to the transmembrane protein typeI family, with a luminal domain containing the carbohydrate recognition domain(CRD), a transmembrane domain and a short C-terminal cytosolic tail. Thesequence of MR60 (Arar et al., 1995) was found to be identical to that of ERGIC-53 (Schindler et al., 1993), a protein of the endoplasmic reticulum (ER) Golgi inter-mediate compartment. The human MR60 protein is homologous to the rat p58protein (Lahtinen et al., 1996). Recently, we have shown that the truncated recombi-nant protein, lacking both the transmembrane domain and the cytosolic tail, keptits capacity to bind mannosides as long as the intraluminal part is conserved andable to form a dimer (Carriere et al., 1999).

Some understanding of the role of this mannose specific lectin in intracellulartraffic has been obtained by using antibodies prepared by a conventional way: mono-clonal antibodies (Gl/93 anti-ERGIC-53) prepared from animals injected with puri-fied Golgi membranes of Caco-2 cells (Schweizer et al., 1988), monoclonal antibodiesobtained after injection of the purified MR60 in mice (Carpentier et al., 1994) andpolyclonal antibodies raised against a luminal peptide of the rat p58 protein (Laht-inen et al., 1996). However, to further study the physiochemical properties and func-tions of MR60, we wanted to have access to a non-limiting amount of monospecificantibodies and therefore we undertook a program of DNA immunization. Severalexcellent reviews discuss the advantages and the limitations of DNA immunizationand the present state of knowledge on the mechanism by which DNA stimulates theimmune system (see for instance, Ramsay et al., 1997; Tighe et al., 1998). It hasbeen recently reported that cellular localization of an antigen may have an influenceon the immune response; for instance, Boyle et al. (1997) have shown that DNAencoding a secreted protein generated 10 to 100 fold higher response than that enco-ding its membrane associated form. So, we decided to compare DNA constructsencoding either the whole protein or a truncated protein in order to select the bestconstruct to routinely prepare the expected monospecific antisera.

The first report on direct gene transfer in vivo describes injections in mouseskeletal muscle cells (Wolff et al., 1990). Later, Tang et al. (1992) used DNA coatedgold microprojectiles for eliciting an immune response. However, muscle is notknown as an efficient site for antigen presentation. On the other hand, skin containsspecialized cells which enhance immune response and this tissue has been proposedas a better candidate for immunization. Raz et al. (1994) have shown that intra-dermal administration of free plasmid DNA induced long-lasting production of anti-bodies and cytotoxic T cells. Subcutaneous injection of a plasmid at the base of thetail yields serum titers as high as those obtained by intramuscular DNA injectioninto cardiotoxin-pretreated muscle (Bohm et al., 1998). The induction of immuneresponse by DNA injection into the skin may be related to the presence of skin-derived denditric cells (Condon et al., 1996; Casares et al., 1997).

Two different expression plasmids—one encoding the full length MR60 mem-brane protein, the other encoding the truncated MR60A473 protein lacking both thetransmembrane and the cytosolic domains—were delivered by intradermal injection

Intradermal DNA Immunization 561

to mice. The presence of specific antibodies in the mice sera was checked by immuno-fluorescence on permeabilized cells and by immunoprecipitation of cell extractsfollowed by gel electrophoresis.

MATERIALS AND METHODS

Reagents

[35S] methionine (37 MBq/mmol) was from Amersham (Buckinghamshire, UK);cell culture media (DMEM, and DMEM without L-methionine), fetal bovine serum(FBS) and methionine from Gibco (Grand Island, NY); dithiothreitol (DTT), pro-tease inhibitor cocktail specific for mammalian cell extracts [4-(2-aminoethyl)-benzenesulfonyl fluoride, pepstatin A, transepoxysuccinyl-L-leucylamido(4-guanido)butane, bestatin, leupeptin, aprotinin], as well as phenylmethylsulphonylfluoride (PMSF), Triton X-100, paraformaldehyde (PFA), saponin, bovine serumalbumin (BSA) and glycine were from Sigma (Saint-Quentin-Fallavier, France).Anti-FLAG BioM2 antibody was from Eastman Kodak Company (New Haven,CT). The rabbit polyclonal anti-luminal p58 peptide antiserum raised against a pep-tide corresponding to amino acids 158-170 from the rat protein (Lahtinen et al.,1996) was kindly provided by U. Lahtinen (Karolinska Institute, Solna, Norway);the mouse monoclonal Gl/93 antibody raised against purified Golgi membranes ofhuman Caco-2 cells (Schweizer et al., 1988) was kindly provided by H. P. Hauri(Biozentrum, Basel, Switzerland). FTC-labeled goat anti-mouse and FTC-labeledgoat anti-rabbit antibodies were from Bio-Rad (Hercules, CA). TRTC-streptavidinewas from Sigma; rabbit anti-mouse immunoglobulin antibody, rabbit non-immuneserum and fixed Staphylococcus aureus cell suspension (Pansorbin) were from Calbi-ochem-Novabiochem (La Jolla, CA). Enlightning™ was from Du Pont de Nemours(Brussels, Belgium).

Recombinant cDNAs

The pMR60 and pMR60A473 expression vectors encoding the full length MR60protein and a truncated (473 amino acid-long) MR60 protein lacking both the trans-membrane and the cytosolic domains, respectively, were obtained as described byCarriere et al. (1999). The pMR60A473 FLAG vector encoding the truncated (473amino acid-long) MR60 protein and carrying a FLAG epitope (Eastman KodakCompany) in a C-terminal position was constructed as follows: PCR was carriedout using the pMR60 vector with the sense oligonucleotide carrying a Bgl II (5'-AGATCT-3') site in 5' extremity and an antisense oligonucleotide carrying an EcoRV (5'-GATATC-3') site in 3' as follows: 5'-CTC AGA TCT AAG ATG GCG GGATCCAAGCAA-3' and also 5'-TGC TGG ATA TCT TGG AAA TGG TGG TAGTTCTGGGCA-3'. The PCR product was digested by Bgl II/EcoR V and insertedinto an intermediate vector pFLAG (a pcDNA3 vector in which the sequence codingthe FLAG epitope has been inserted using PCR, kindly provided by F. Piller, CBM,Orleans, France) opened by BamH I/EcoR V. All PCR products were sequenced

562 Carriere, Landemarre, Altemayer, Motta, Monsigny, and Roche

using the Thermo Sequenase radiolabeled terminator cycle sequencing kit (Bucking-hamshire, UK), confirming that no error was introduced by PCR.

Immunization of Mice

The plasmids pMR60A473 and pMR60, used for mice immunization, were pre-pared from Escherichia coli by centrifugation in a cesium chloride gradient. FemaleBalb/C mice, about 5 weeks old, were injected with the plasmids suspended in nor-mal saline at a concentration of 1 mg/ml. Five mice were intradermally injected atthe basis of the tail using a 1 ml syringe and a 0.4 x 12 mm needle (Sherwood Medi-cal, Evry, France) with 100 ug plasmid on days 0, 7, 28 and 42. Blood was collectedone day before starting the immunization and on days 35, 49 and 91. The sera werestored at 4°C in the presence of 0.1 mg/ml sodium azide.

Cell Culture and Transfections

Cos-7 and HeLa cells (ATCC, Rockville, MD) were grown in DMEM sup-plemented with 10% fetal bovine serum (FBS), 100 I.U./ml penicillin and 100 ug/mlstreptomycin. Cells were transfected by using a polyethyleneimine (PEI 25,000,10 mM) protocol according to Boussif et al. (1995). For metabolic labeling experi-ments, cells (2.106) were plated in dishes, 10 cm in diameter. 24 hr before transfec-tion; the complexes were formed by mixing 15 ug of the pMR60 DNA and 22.5 ulof PEI. For immunofluorescence microscopy analysis, cells (4.104) were plated oneday before transfection on glass slides in four-well multichambers and transientlytransfected with a complex formed with 5 ug DNA and 7.5 u1 PEI. Biosyntheticlabeling as well as immunofluorescence analysis were performed 24 hr aftertransfection.

Immunofluorescence Microscopy

All incubations with serum, antiserum or antibody solution were conducted atroom temperature. Untreated cells or cells transiently transfected with eitherpMR60A473 or pMR60A473 FLAG were washed twice with phosphate-buffered saline(PBS) at pH 7.4, fixed for 30 min in PBS containing 20 mg/ml paraformaldehyde(PFA), washed twice with PBS containing 20 mM glycine and permeabilized 20 minin PBS containing 1 mg/ml saponin and 20 mM glycine. Before adding either theanti-luminal p58 peptide antiserum, the anti-MR60 antiserum or the anti-MR60A473

antiserum, fixed cells were incubated for 30 min in PBS containing 2% (per volume)goat serum in order to prevent unspecific binding (Lahtinen et al., 1996) and thenfurther incubated for 45 min with either the anti-luminal p58 peptide antiserum at a1/300 dilution, the anti-MR60 or the anti-MR60A473 antiserum at a 1/500 dilutionor the monoclonal Gl/93 antibody at a 1/300 dilution in PBS containing 1 mg/mlsaponin. Cells were then washed four times in PBS containing 1 mg/ml saponin andincubated 30 min with either FTC-labeled goat anti-mouse or with FTC-labeled goatanti-rabbit antibodies. For double immunofluorescence experiments, the fixed/permeabilized transfected cells were sequentially incubated with mouse anti-MR60

Intradermal DNA Immunization 563

antiserum, FTC-labeled goat anti-mouse antibodies, anti-FLAG BioM2 antibody(at a 1/300 dilution) and TRTC-labeled streptavidin (at a 1/100 dilution). Afterextensive washes (four times), cells were mounted in a PBS-glycerol mixture (50%per volume) containing l0 mg/ml DABCO (l,4-diazabicyclo[2,2,2]octane) used asan antifading agent (Johnson et al., 1982).

Confocal Microscopy Analysis

Cells were analyzed with a confocal imaging system MRC-1024 (Hercules, CA)equipped with a Nikon microscope (Nikon, Tokyo, Japan), a 60 x Planapo objective(numerical aperture 1.4) and a krypton/argon laser tuned to produce both 488-nmfluorescein excitation and 564-nm rhodamine excitation wavelength beams allowingsimultaneous reading of both fluorescent signals. Pictures were recorded with a kal-man filter (average of 10 to 15 images). Images were treated using an Adobe Photo-shop software (Adobe Systems Inc., San Jose, CA).

Metabolic Labeling and Immunoprecipitation

For radiolabeling, cells at 50% confluency were incubated (30 min at 37°C) inmethionine-free medium containing 10% dialyzed FBS before adding [35S] methion-ine (9 MBq per 10-cm dish) for 30 min at 37°C. After labeling, cells were chased(1 hr 30 at 37°C) in a complete culture medium containing 2 mM unlabeled methion-ine. Cells were then washed twice in PBS, harvested, and lysed in the lysis buffer(150 mM NaCl, l0 mM Tris-HCl, pH 7.4 containing 1% Triton X-100, the cocktailof protease inhibitors specific for mammalian cells (100 ul/107 cells) and 1 mMPMSF). The lysates were cleared by centrifugation (15,000g for 15 min), and thesupernatants were diluted in 1 ml of dilution buffer (5 mM EDTA, 250 mM NaCl,20 mM Tris-HCl at pH 8 containing 1% Triton X-100 and 2.5 mg/ml BSA). Beforeimmunoprecipitation of the endogenous MR60, the diluted Cos cell lysate wascleared by an overnight incubation at 4°C with 10 ul non-immune rabbit serumfollowed by an addition of 10% suspension of fixed Staphylococcus aureus cells(30 min at 4°C) as described by Carlsson and Fukuda (1986). After centrifugation(30 s at 10,000g), the immunoprecipitation of the endogenous MR60 was performedby adding either 5 ul of the anti-luminal p58 peptide antiserum or 5 u1 of the anti-MR60 antiserum to the supernatant followed by an overnight incubation at 4°C.Immune complexes were harvested as indicated above after addition of fixed S.aureus cells. The immune complexes were then washed once with the dilution buffer,three times with a high salt buffer (5 mM EDTA, 400 mM NaCl, 20 mM Tris-HClat pH 8 containing 1% Triton X-100) and once with 10 mM Tris-HCl at pH 7.4. Theimmunoprecipitated proteins were solubilized at 100°C in the loading buffer (60 mg/ml sodium dodecyl sulfate (SDS), 375 mM Tris-HCl, at pH 6.8 containing 30% gly-cerol (v/v) and 100 mM dithiothreitol (DTT) used as a reducing agent. Proteins wereseparated in a 1 mg/ml SDS-10% polyacrylamide slab gel. Proteins were fixed in amixture of isopropanol, acetic acid and water (25:10:65, per volume) and treatedfor fluorography with Enlightning™ according to the manufacturer's instructions.

564 Carriere, Landemarre, Altemayer, Motta, Monsigny, and Roche

RESULTS AND DISCUSSION

Two expression vectors (Fig. 1), one pMR60 expressing the full length MR60protein with its transmembrane segment and its cytosolic tail, and the other onepMR60A473 encoding the truncated MR60M473 protein, lacking the C-terminal partcorresponding to the transmembrane and cytosolic domains have been constructed,to perform DNA immunization.

We have previously shown that these two vectors (bearing or not a myc epitopeon their C-terminal end), transfected into eukaryotic cells allow the expression ofrecombinant proteins that can be either visualized into cells or immunoprecipitatedusing anti-myc or anti-p58 antibodies (Carriere et al., 1999). The full length MR60recombinant protein, as well as the endogenous MR60 protein, spontaneously oligo-merize in dimeric and hexameric forms after synthesis. The truncated recombinantprotein MR60M473 formed dimers, but not hexamers, while it kept its sugar bindingactivity (Carriere et al., 1999). The homologous rat p58 protein, in which one of thetwo cysteines close to the transmembrane region had been mutated, was also shownto form dimers and to keep its ability to bind mannoside-substituted beads (Lahtinenet al., 1999).

The two expression vectors (without a myc C-terminal tag) have been usedsuccessfully to immunize mice by intradermal injection. In preliminary experiments,the plasmids injected into quadriceps muscles led either to no detectable immuneresponse in five mice in the case of the cDNA encoding the full length protein, oronly to a low antibody production in two out of five mice in the case of thepMR60A473 vector. We then tested an intradermal protocol. Before immunization,the serum of each mouse was tested; the mice, of which the serum gave a highintracellular fluorescent background, were not used for immunization. Mice (five ineach experiment) received three injections of 100 ug plasmid before the first bleeding.Mice were boosted at days 7, 28 and 42. Sera were collected once a week and testedon fixed and permeabilized cells to look for the binding of antibodies specific forthe endogenous MR60 protein. The characteristic cellular location of the MR60 andp58 proteins is known (Schweizer et al., 1988; Saraste and Svensson, 1991): theseproteins are mainly located in the proximity of the nucleus in the Golgi region. Inpermeabilized Cos cells, sera obtained from mice immunized either with the full

Fig. 1. Scheme of the endogenous and recombinant MR60 proteins. MR60 protein is a510 amino acid-long protein containing a luminal domain (D), a putative transmembranedomain (D) and a cytosolic tail (•). The MR604473 protein (473 amino acids) correspondsto the MR60 protein lacking both the transmembrane and the cytosolic domains. ThepMR60A473 FLAG contains a FLAG (0) peptide at its C-terminal end.

Intradermal DNA Immunization 565

Fig. 2. Immunofluorescence staining of the endogenous MR60 in Coscells. Cos cells were fixed, permeabilized and then incubated with rab-bit anti-luminal p58 peptide antiserum raised against a peptide of therat p58 protein (a), with the Gl/93 monoclonal antibody specific forthe ERGIC-53 protein (b), with anti-MR60A473 antiserum (c) obtainedby DNA immunization using the pMR604473 cDNA, or with anti-MR60 antiserum (d) obtained by DNA immunization using thepMR60 cDNA. FTC-goat anti-rabbit (a) or FTC-goat anti-mouse (b),(c) and (d) were used to visualize the first antibodies. The bar corre-sponds to 15 um in (a) and (b), and 10 um in (c) and (d).

length cDNA (anti-MR60 antiserum, Fig. 2(d)) or with the truncated pMR60A473

cDNA (anti-MR60M473 antiserum, Fig. 2(c)) gave similar patterns which resemblethose obtained with the monoclonal anti-ERGIC-53 (Fig. 2(b)) or the anti-p58 pep-tide serum (Fig. 2(a)), i.e. the immunofluorescence comes from vesicular elementsconcentrated in the Golgi region in addition to some fluorescence spread throughoutthe cytoplasm. Four out of the five mice immunized with the truncated form gavehigh titer antibody; by flow cytometry analysis the labeling with a 1/4000 dilutionof the whole serum was still strong (data not shown). In contrast, only one out ofthe five mice, immunized with the pMR60 vectors encoding the full length protein,gave a high titer specific antibody, the sera of the other mice were negative evenwhen they were used at 1/100 dilutions.

Both the anti-MR60 and anti-MR60A473 antisera allowed the intracellular label-ing of the endogenous MR60 protein as well as the recombinant MR60 proteins.

566 Carriere, Landemarre, Altemayer, Motta, Monsigny, and Roche

Fig. 3. Double immunofluorescence staining of recombinant MR604473

FLAG protein in Cos cells. Cos cells transfected with the pMR60A473 FLAGcDNA were fixed 24 hr after transfection, permeabilized and then incubatedsequentially with anti-MR60A473 antiserum, FTC-goat anti-mouse, anti-FLAG BioM2 antibody and TRTC-streptavidin. The staining of the FLAGepitope was analyzed through the rhodamine channel (a) and the staining ofboth the endogenous and the recombinant proteins through the fluoresceinchannel (b); bar: 10 uM.

Fig. 4. Immunofluorescence staining of the endogenous and recombi-nant MR60 in HeLa cells using the anti-MR60 antiserum. HeLa cellsor HeLa cells transfected with the pMR60a473 cDNA were fixed, perme-abilized and then incubated with the anti-MR60 antiserum followed byFTC-goat anti-mouse to stain the endogenous MR60 protein (a) or boththe endogenous and the recombinant proteins (b); bar 10 um.

Intradermal DNA Immunization 567

Fig. 5. Specificity of the anti-MR60A473 antiserum. Immunoprecip-itates were obtained with anti-peptide p58 antiserum (lane 1) orwith anti-MR60A473 antiserum (lane 2) from [35S]-methionine-labeled Cos cells: both antisera recognize a 57 kDa protein corre-sponding to the endogenous MR60 protein.

Double immunofluorescence labeling, performed on Cos cells transfected withpMR60A473 FLAG using on the one hand the anti-MR60A473 antiserum (Fig. 3(b))and on the other hand the anti-FLAG antibody (Fig. 3(a)) showed an extendedoverlapping pattern. Single fluorescence staining of HeLa cells and HeLa cellstransfected with the plasmid encoding the truncated MR60A473 protein with theserum of the mouse immunized with the full length cDNA showed a labeling charac-teristic of the endogenous (Fig. 4(a)) MR60 and of the recombinant MR60A473

(Fig. 4(b)) proteins, respectively. Immunoprecipitation experiments were carried outto further characterize the quality of the antibodies. The immunoprecipitation of theendogenous MR60 lectin with either the anti-p58 peptide serum or the anti-MR60A473 serum was performed using [35S] methionine labeled Cos cells. As shownin Fig. 5 (lane 2), the anti-MR60A473 antiserum specifically precipitated a proteinwhich, in electrophoresis under reducing conditions, gives a 57,000 Mr band corre-sponding to the endogenous MR60 protein. The pattern obtained upon immunopre-cipitation of the endogenous MR60 using the anti-p58 peptide antiserum (lane 1)was quite similar. Therefore, the anti-MR60A473 antiserum obtained after DNAimmunization by expressing the vector pMR60A473 is quite specific of the MR60lectin.

In conclusion, highly specific antibodies against membrane proteins wereobtained by DNA immunization. The main findings are as follows: a cDNA enco-ding a truncated protein, lacking the transmembrane domain, was more efficientthan the cDNA encoding the full length protein. After intramuscular injection ofthe cDNA encoding the full length protein, no antibody was detected; however,after intradermal injection of the same plasmid, high specific antibody response wasobtained in one out of five mice. In contrast, after intradermal injection of a cDNAencoding a protein deleted of both the transmembrane and cytosolic domains, theantisera had high titer in four out of five mice. Antisera of all positive mice werevaluable for in situ visualization as well as for immunoprecipitation. The quality ofthe labeling and that of the electrophoretic pattern of immunoprecipitated materialobtained by using antisera from cDNA immunized mice were as good as or betterthan those obtained with an anti-ERGIC-53 monoclonal antibody or an anti-peptidepolyclonal antiserum.

568 Carriere, Landemarre, Altemayer, Motta, Monsigny, and Roche

ACKNOWLEDGEMENTS

We thank Francoise Fargette for her valuable help in producing the variousplasmids and Marie-Therese Bousser for her skilled technical assistance in cellexperiments. We are grateful to Dr. Ulla Lahtinen (Ludwig Institute for CancerResearch, Stockholm) for providing us with the p58 anti-luminal peptide antiserumand to Dr. Hans-Peter Hauri (Biozentrum, Basel, Switzerland) for providing us withthe Gl/93 monoclonal antibody. V.C. received a fellowship from the Ministere del' Enseignement Superieur et de la Recherche Scientifique et Technique. This work wassupported by grants from the Agence Nationale de Recherche sur le SID A and fromthe Association pour la Recherche sur le Cancer (ARC 6132). L.L. is Research Labor-atory Director at Agro-Bio SA (La Ferte Saint Aubin, France). M.M. and V.A. areProfessor and Assistant-Professor, respectively, at the Universite d' Orleans andA.C.R. is Research Director at the Institut National de la Sante et de la RechercheMedicale.

REFERENCES

Arar, C., Carpentier, V., Le Caer, J. P., Monsigny, M., Legrand, A., and Roche, A. C. (1995) ERGIC-53, a membrane protein of the endoplasmic reticulum-Golgi intermediate compartment, is identicalto MR60, an intracellular mannose-specific lectin of myelomonocytic cells. J. Biol. Chem. 270:3551-3553.

Bohm, W., Mertens, T., Schirmbeck, R., and Reimann, J. (1998) Routes of plasmid DNA vaccinationthat prime murine humoral and cellular immune responses. Vaccine 16:949-954.

Boussif, O. et al. (1995) A versatile vector for gene and oligonucleotide transfer into cells in culture andin vivo: polyethyleneimine. Proc. Natl. Acad. Sci. (USA) 92:7297-7301.

Boyle, J. S., Koniaras, C., and Lew, A. M. (1997) Influence of cellular location of expressed antigen onthe efficacy of DNA vaccination: cytotoxic T lymphocyte and antibody responses are suboptimalwhen antigen is cytoplasmic after intramuscular DNA immunization. International Immun. 9:1897-1906.

Carlsson, S. R., and Fukuda, M. (1986) Isolation and characterization of leukosialin, a major sialoglyco-protein on human leukocytes. J. Biol. Chem. 261:12779-12786.

Carpentier, V., Vassard, C., Plessis, C., Motta, G., Monsigny, M., and Roche, A. C. (1994) Characteriz-ation and cellular localization by monoclonal antibodies of the 60 kDa mannose specific lectin ofhuman promyelocytic cells, HL60. Glycoconjugate J. 11:333 -338.

Carriere, V., Piller, V., Legrand, A., Monsigny, M., and Roche, A. C. (1999) The sugar binding activityof MR60, a mannose-specific shuttling lectin requires a dimeric state. Glycobiology (in press).

Casares, S., Inaba, K., Brumeanu, T. D., Steinman, R. M., and Bona, C. A. (1997) Antigen presentationby dendritic cells after immunization with DNA encoding a major histocompatibility complex classII-restricted viral epitope. J. Exp. Med. 186:1481-1486.

Condon, C., Watkins, S. C., Celluzzi, C. M., Thompson, K., and Falo. L. D. (1996) DNA-based immun-ization by in vivo transfection of dendritic cells. Nature Med. 2:1122-1128.

Johnson, G. D., Davidson, R. S., McNamee, K. C., Russell, G., Goodwin, D., and Holborow, E. J.(1982) Fading of immunofluorescence during microscopy: a study of the phenomenon and its remedy.J. Immunol. Methods 55:231-242.

Lahtinen, U., Hellman, U., Wernstedt, C., Saraste, J., and Pettersson, R. F. (1996) Molecular cloningand expression of a 58-kDa cis-Golgi and intermediate compartment protein. J. Biol. Chem.271:4031-4037.

Lahtinen, U., Svensson, R., and Pettersson, R. F. (1999) Mapping of structural determinants for theoligomerization of p58, a lectin-like protein of the intermediate compartment and cis-Golgi. Eur. J.Biochem. 260:392-397.

Intradermal DNA Immunization 569

Pimpaneau, V., Midoux, P., Monsigny, M., and Roche, A. C. (1991) Characterization and isolation ofan intracellular D-mannose-specific receptor from human promyelocytic HL60 cells. Carbohydr. Res.213:95-108.

Ramsay, A. J., Ramshaw, I. A., and Ada, G. L. (1997) DNA immunization. Immunol. Cell Biol. 75:360-363.

Raz, E. et al. (1991) Distribution of the intermediate elements operating in ER to Golgi transport. J.Cell Science 100:415-430.

Schindler, R., Itin, C., Zerial, M., Lottspeich, F., and Hauri, H. P. (1993) ERGIC-53, a membraneprotein of the ER-Golgi intermediate compartment, carries an ER retention motif. Eur. J. Cell Biol.61:1-9.

Schweizer, A., Fransen, J. A., Bachi, T., Ginsel, L., and Hauri, H. P. (1988) Identification, by a mono-clonal antibody, of a 53-kD protein associated with a tubulo-vesicular compartment at the cis-sideof the Golgi apparatus. J. Cell Biol. 107:1643-1653.

Tang, D. C., De Vit, M., and Johnston, S. A. (1992) DNA immunization is a simple method for elicitingan immune response. Nature 356:152-154.

Tighe, H., Corr, M., Roman, M., and Raz, E. (1998) Gene vaccination: plasmid DNA is more than justa blueprint. Immunol. Today 19:89-97.

Wolff, J. A. et al. (1990) Direct gene transfer into mouse muscle in vivo. Science 247:1465-1468.