The transforming rat Her-2/neu oncogene embedded into the genome of virgin transgenic BALB/c mice (BALB-neuT) provokes the development of an invasive carcinoma in each of their 10 mammary glands. i.m. vaccination with DNA plasmids coding for the extracellular and transmembrane domains of the protein product of the Her-2/neu oncogene started when mice already display multifocal in situ carcinomas temporarily halts neoplastic progression, but all mice develop a tumor by week 43. By contrast, progressive clearance of neoplastic lesions and complete protection of all 1-year-old mice are achieved when the same plasmids are electroporated at 10-week intervals. Pathological findings, in vitro tests, and the results from the immunization of both IFN-γ and immunoglobulin gene knockout BALB-neuT mice, and of adoptive transfer experiments, all suggest that tumor clearance rests on the combination of antibodies and IFN-γ-releasing T cells. These findings show that an appropriate vaccine effectively inhibits the progression of multifocal preneoplastic lesions.
The concept of the preventive efficacy of antitumor vaccines comes from a multitude of experiments showing that preimmunization efficiently inhibits subsequent lethal tumor challenges (1 , 2) . There is overwhelming evidence that eradication of fast growing transplantable tumors rests on swift high-affinity T-cell reactivity (3) , whereas B cells may interfere with the efficiency of the reaction (4) . By contrast, the question of whether a vaccine can inhibit the progression of carcinogenesis is rarely addressed (5) . Although several theoretical considerations suggest that the chronic progression of early neoplastic lesions is an appropriate target for an immune control (6) , the effective inhibitory potential of T cell and antibody-mediated reaction mechanisms elicited by a vaccine is still undetermined. This assessment is important because it may provide new information which may not coincide with that obtained from the study of fast growing transplantable tumors (7) . Progression of a preneoplastic lesion is a lengthy process that may be hampered by mechanisms that are not efficacious when confronted with the unnatural speed of transplantable tumors.
Tumors which develop in transgenic mice as a consequence of a defined gene alteration form an experimental model for evaluation of the preventive efficacy of innate (8) and adaptative (7 , 9) immunity. In these mice, tumors become evident after progressive stages of tumorigenesis, and the relationship between the tumor and surrounding tissues is preserved (10) . Despite these important features, transgenic mouse models of cancer are not devoid of pitfalls. The transgene may not follow the same developmental expression pattern as the natural gene, because its promoter’s own peculiarities may not be shared by the mechanisms involved in the pathogenesis of natural tumors. The timing of transgene first expression during ontogenesis has a crucial importance because it shapes the features of mouse immune tolerance to the transgene protein product.
Thanks to the work of Leder and Muller (11) , there are several lines of mice transgenic for the wild-type or transforming (T) rat (r) Her-2/neu oncogene under the control of mouse mammary tumor virus promoter. rp185neu shares 94.8% sequence homology with its mouse counterpart (12) . The females of the distinct transgenic lines overexpress rp185neu at different periods of their life and develop multiple Her-2/neu mammary carcinomas after varying periods of latency (8) . BALB/c female mice made transgenic for the rHer-2/neuT display one of the most aggressive progressions of Her-2/neu carcinogenesis (8) . rp185neu is already overexpressed on the surface of the cells of the rudimentary mammary (10) , salivary (13) , and Harderian (data not shown) glands of 3–4-week-old mice. At 6 weeks, rp185neu cells give rise to a widespread mammary atypical hyperplasia, which progresses to form an invasive and metastasizing carcinoma that becomes palpable in all 10 mammary glands between the 22nd and 27th week of age (8) . No reactive lymphocyte infiltration nor antibody response is associated with carcinogenesis progression (7 , 8) .
In wild-type BALB/c mice, DNA vaccination with plasmids coding for the extracellular (EC) and transmembrane (TM) domains of rp185neu (p185EC-TM plasmids) triggers a strong antibody and CTL response, protects all mice against a lethal challenge by a transplantable rp185neu+ carcinoma, and even leads to the eradication of clinically evident rp185neu+ tumors (7 , 14) . As a consequence of the early and diffuse tissue overexpression of endogenous rp185neu, the same DNA plasmids fail to trigger in BALB-neuT mice the strong and rapid T-cell reactivity required for an efficient rejection of challenging rp185neu+ carcinoma cells.
No CTL and a poor antibody response only is elicited (7) . Induction of an immunity able to protect against Her-2/neu carcinogenesis is thus a daunting challenge (15) . Nevertheless, in the present study, we show that the reactivity elicited through electroporation of p185EC-TM plasmids leads to a progressive and sustained clearance of already present multifocal preneoplastic lesions in all mammary glands and keeps all BALB-neuT mice free from palpable tumors at 1 year of age. No protection takes place in the absence of IFN-γ and antibody. Adoptive transfer experiments show that protection rests on both antibody and T cell-mediated reactivity, even if a CTL activity is not particularly evident.
MATERIALS AND METHODS
Virgin BALB-neuT female mice (H-2d) overexpressing the transforming rHer-2/neu oncogene under the control of the mouse mammary tumor virus (8) were bred by us. BALB-neuT mice knocked out for the IFN-γ gene (BALB-neuT/γKO) and those knocked out for the immunoglobulin μ chain gene (BALB-neuT/μKO) were generated by crossing BALB-neuT mice with BALB/c mice KO for IFN-γ gene from The Jackson Laboratory (Bar Harbor, ME) and BALB/c mice KO for the immunoglobulin μ chain kindly provided by Dr. T. Blankenstein (Berlin, Germany; Ref. 4 ), respectively. Mice were treated according to the European Community guidelines. Female transgenic BALB-neuT mice of the same age were randomly assigned to the control and treatment groups, and all groups were specifically treated concurrently. As each experiment was repeated two to four times with similar results, data were cumulated and reported in the figures. Mammary glands were palpated at weekly intervals to note tumor appearance. Progressively growing masses > 1 mm mean diameter were regarded as tumors. Tumor multiplicity was calculated as the cumulative number of incident tumors per total number of mice, and it is reported as mean ± SE (8) .
rp185neu+ TUBO cells are from a carcinoma arising in a BALB-neuT mouse (7) . N202.1A (rp185neu+) and N202.1E (rp185neu−) lines were taken randomly from a carcinoma of a FVB mouse (H-2q) transgenic for the rat Her-2/neu protoncogene (16) . Cells were cultured in DMEM (BioWhittaker Europe, Verviers, Belgium) with 20% FBS (Life Technologies, Inc., San Giuliano, Italy).
Injection of p185EC-TM Plasmids and Electric-Pulse Delivery.
pcDNA3 vector coding the EC and TM domains of rp185neu was produced and used as described (7) . Briefly, DNA was precipitated, suspended in sterile saline at 1 mg/ml, and stored in aliquots at −20°C for use in immunization protocols. A total of 100 μl of this solution (100 μg of DNA) was injected into the surgically exposed quadriceps of anesthetized mice at weeks 10 and 12. For DNA electroporation, 25 μg of p185EC-TM plasmids in 20 μl of 0.9% of NaCl with 6 mg/ml polyglutamate were injected into the tibial muscle of anesthetized mice at weeks 10 and 12. Electric pulses were applied by two electrodes placed on the shaved skin covered with a conducting gel. Two square-wave 25 ms, 375 V/cm pulses were generated by a T820 electroporator (BTX, San Diego, CA). Each course of DNA i.m. vaccination or DNA electroporation consisted of two administrations with an interval of 14 days.
Whole Mount Image Analyses.
Whole mounts of all mammary glands were performed as indicated on the Internet. 6 Images were acquired by dividing each whole mount into 10 quadrants. Ten points were randomly chosen on the duct surface in each quadrant, and the corresponding lesions were measured. All lesions on a quadrant with a diameter > 150 μm were counted. Pictures were taken with a Nikon Coolpix 950 digital camera (Nital S.p.A., Torino, Italy) mounted on a stereoscopic microscope (MZ6; Leica, Milano, Italy) with a 0.63 objective giving a total magnification of ×6.3. The resolution was 1600 × 1200 pixels. Images were acquired with an Adobe PhotoShop v 6.0 graphic software (Adobe Systems; San Jose, CA). Mammary glands that exceeded the size of a single imaging area were captured by photographing contiguous fields in a raster pattern. Each captured image was merged using the layer technique in Adobe PhotoShop to form a single composite picture for analysis. Spatial calibration was determined by photographing a 1-mm stage using the same parameters as those for image capturing of whole mount preparations. The distance drawn on the 1-mm calibration image was divided by 1000 to find the number of pixels per micrometer. In each image, ≥100 discrete points were randomly chosen on the duct surface, the width in micrometers of any lesion in that point was measured perpendicular to the duct direction, and a value of zero was assigned if no lesion was present. Measurements for all lesions were recorded, and mean and SE were calculated for each treatment group. For each mammary gland, the number of lesions with a diameter > 150 μm was also recorded.
Groups of three mice were sacrificed at the indicated times, and mammary tissue was processed for morphological analysis (10) . For histological evaluation, tissue samples were fixed in 10% neutral buffered formalin, embedded in paraffin, sectioned at 4 μm, and stained with H&E. For immunohistochemistry, acetone-fixed cryostat sections were incubated for 30 min with antidendritic cell antibodies (NLDC 145; Cederlane Laboratories, Ontario, Canada), anti-CD11c (Chemicon International, Inc., Temecula, CA), anti-CD4, anti-CD8a (both from Sera-Lab, Crawley Down, Sussex, United Kingdom), anti-Mac1 (anti-CD11b/CD18), anti-Mac3 (both from Boehringer Mannheim, Milan, Italy), antigranulocytes (RB6–8C5; provided by Dr. R. L. Coffman, DNAX, Inc., Palo Alto, CA), anti-interleukin1β (Genzyme, Cambridge, MA), and anti-IFN-γ mAb (provided by Dr. S. Landolfo, Torino University, Torino, Italy). To evaluate the expression of rp185neu and proliferating cell nuclear antigen, sections were incubated with polyclonal rabbit antineu antibody (C-18; Santa Cruz Biotechnology, Santa Cruz, CA) and antiproliferating cell nuclear antigen (Ylem, Roma, Italy) antibody, overlaid with biotinylated goat antirat, antihamster, and antirabbit or horse antigoat immunoglobulin (Vector Laboratories, Burlingame, CA) for 30 min and incubated with avidin-biotin complex/AP (DAKO, Glostrup, Denmark). At 52 weeks of age, the heart, kidney, and liver of each mouse were examined. Morphological studies were conducted by three pathologists in a blind fashion.
In Vitro Assays.
In all tests, spleen cells (Spc) obtained from mice in each treatment group were depleted of erythrocytes by hypotonic lysis and cultured in RPMI 1640 supplemented with 10% fetal bovine serum. To evaluate IFN-γ production by fresh T cells, Spc were stimulated with anti-CD28 and -CD3 (1 μg/ml final concentration; PharMingen, San Diego, CA) for 8 h. IFN-γ-producing cells were identified using the mouse IFN-γ cell enrichment and detection kit (Miltenyi Biotec, Bergisch-Gladbach, Germany). Activated Spc were labeled with an anti-IFN-γ (clone R4–6A2) conjugated with an anti-CD45 (clone 30S11) monoclonal antibody (mAb; Miltenyi Biotec) for 5 min on ice, then incubated for 45 min at 37°. Cross-staining was avoided by keeping the cell density at 1 × 105 cells/ml. IFN-γ bound to the capture matrix was stained with phycoerythrin (PE)-conjugated mAb against IFN-γ (clone AN.18.17.24; Miltenyi Biotec). Anti-PE microbeads were used to enrich PE (IFN-γ)-stained cells with two rounds on a magnetic separator (MS+ MACS; Miltenyi Biotec).
The cells were counterstained with mAb against CD8-FITC (PharMingen) and analyzed by flow cytometry. To evaluate CTL activity, 1 × 107 Spc were stimulated for 6 days in a mixed lymphocyte tumor interaction (MLTI) with 5 × 105 mitomycin-C (Sigma, St. Louis, MO)-treated TUBO cells in the presence of 10 units/ml interleukin-2 (Eurocetus, Milan, Italy). CTL activity was then assayed in a 48-h [3H]TdR release assay at E:T TUBO cells ratio from 50:1 to 6:1 in round bottomed 96-well microtiter plates in triplicate. The results were then expressed as LU20/107 effector cells (17) . To quantify IFN-γ production after MLTI, Spc (1 × 105) were cocultured in RPMI 1640 supplemented with 10% fetal bovine serum in the presence of 1 × 105 TUBO cells. Supernatants were collected after 24 h and tested with a mouse IFN-γ ELISA kit (PharMingen) according to the manufacturer’s protocol.
Spc were stained immediately or after MLTI restimulation with mAb against CD4 or CD8a (Cedarlane, Hornby, Ontario, Canada). Intracellular cytokine assay was performed after polyclonal activation with the Leukocytes Activation Cocktail (BD PharMingen) and stained using the Intracellular Cytokine Staining Starter kit-mouse (BD PharMingen) according to the manufacturer’s recommendations. Samples were analyzed with a FACScan (Becton Dickinson, Mountain View, CA). Data elaborated through CellQuest (Becton Dickinson) software were displayed as a percentage of positivity.
Sera were collected from mice from each treatment group and diluted 1:100. Their binding to rp185neu+ N202.1A and rp185neu− N202.1E cells was assayed by flow cytometry (9) . Isotype determinations were carried out by an indirect immunofluorescence procedure. Dilutions (1:20) of sera in PBS-azide-BSA were incubated with 2 × 105 N202.1A or N202.1E cells for 45 min at 4°C. After washing, the cells were incubated for 30 min with rat biotin-conjugated antibodies antimouse IgA, IgM, IgG1, IgG2a, IgG2b, and IgG3 (Caltag Laboratories, Burlingame, CA), and then for 30 min with 5 μl of streptavidin-phycoeritrin (DAKO), and resuspended in PBS-azide-BSA containing 1 mg/ml propidium iodide and evaluated in a FACScan. The specific N202.1A-binding potential was calculated as follows: [(% positive cells with test serum)(fluorescence mean)] − [(% positive cells with control serum)(fluorescence mean)] × serum dilution (9) . Viable cells (1 × 104) were analyzed in each evaluation.
Antibody-Dependent Cellular Cytotoxicity.
[3H]TdR TUBO cells (5 × 103) were incubated for 2 h at 4°C with progressive dilutions (1:10 to 1:100) in HBSS of serum and washed. Spc from untreated mice were then added at 50:1 E:T ratio to each well in triplicate, and lysis was determined as described (18) .
CD90+ Spc were isolated by magnetic cell sorting with an autoMACS (Miltenyi Biotech). Briefly, Spc were labeled with a PE-conjugated anti-CD90 mAb (clone 53–2.1; PharMingen), and then anti-PE microbeads (Miltenyi Biotech) were used to isolate CD90+ cells by two rounds on a magnetic separator (MS+ MACS; Miltenyi Biotech). CD90+ Spc (1 × 107, 90–92% positive) were administered i.v. in 0.2 ml of PBS to recipient BALB-neuT mice at weeks 10, 12, and 13. Other mice received 0.5 ml of pooled sera from 15-week-old electroporated donors. Lastly, one group received both spleen T cells and sera administered 4 h apart. Recipient BALB-neuT mice were sacrificed at week 17, and the whole mounts of all 10 mammary glands were evaluated. These results were compared with the staging of the fourth right and left mammary glands.
Differences in tumor incidence were evaluated with the Mantel-Haenszel Log-rank test; differences in tumor multiplicity, mammary lesions, LU20, cytokine production, and antibody titer were evaluated with Student’s t test.
DNA Electroporation Inhibits Carcinogenesis.
i.m. DNA vaccination with p185EC-TM plasmids begun when mammary glands display atypical hyperplasia (week 6) impaired carcinogenic progression (7 , 9) . To test whether it was also able to inhibit the progression of more advanced forms, mice received a course of two p185EC-TM plasmids injected i.m. (DNA vaccination course), or electroporated (DNA electroporation course), at the 10th and 12th week of age, when multiple in situ carcinomas were present. A delayed occurrence of the first palpable tumor was observed in a few vaccinated mice, although by week 43, all mice displayed a palpable tumor (Fig. 1A) ⇓ . The protection was not significantly improved by simply repeating the vaccination courses at weeks 20–22 in those few mice in which tumors were not yet palpable (data not shown). By contrast, 47% of DNA-electroporated mice remained tumor free at 1 year of age when the experiment was ended (Fig. 1C) ⇓ . Tumor multiplicity, too, was significantly reduced (Fig. 1C) ⇓ . Two additional DNA electroporation courses at weeks 20–22 and 30–32 extended the tumor-free survival, and no tumor was palpable until week 45. Although the final number of mice without tumor was only marginally extended, tumor multiplicity was significantly lower. Finally, all mice that received four DNA electroporation courses at weeks 10–12, 20–22, 30–32, and 40–42 were tumor free at 1 year of age. Their disease-free survival was more than doubled compared with mice electroporated with the empty vector. The aggressiveness and ineluctability of the progression of the multifocal carcinoma in situ of BALB-neuT mice make this complete and persistent inhibition of advanced preneoplastic lesions a significant finding.
Clearance of in Situ Carcinomas.
In DNA-electroporated mice, the carcinoma in situ lesions present at the time of vaccination (9) were almost unchanged at week 15 (Fig. 2C) ⇓ but markedly reduced by week 21 (Fig. 2D) ⇓ . At week 52, neoplastic side buds had completely vanished in mice receiving the four electroporation courses (Fig. 2, E and F) ⇓ . This progressive clearance of in situ carcinomas in DNA-electroporated mice is associated with a marked decrease of membrane expression of rp185neu. At week 21, rp185neu was only expressed in the cytoplasm, but even there, its expression was fainter than in untreated BALB-neuT mice (Fig. 3C) ⇓ . The poor expression of rp185neu goes along with a marked reduction of epithelial cell proliferation as assessed by antiproliferating cell nuclear antigen staining (Fig. 3F) ⇓ . On the other hand, although numerous dendritic cells (NLDC 145+, CD11c+), macrophages (Mac1+, Mac3+), and CD4 (data not shown) were present in the stroma surrounding the hyperplasia foci, several CD8 (Fig. 3H) ⇓ neutrophils (RB6–8C5+; data not shown) overcame the basal membrane and were briskly intermingled with the residual neoplastic cells present in DNA-electroporated mice. Anti-IFN-γ (Fig. 3I) ⇓ widely stained the cytoplasm of most infiltrating T cells, whereas anti-interleukin-1β antibody stained areas rich in infiltrating macrophages (data not shown). At week 52, rp185neu+ cells had vanished in the mice that received the four electroporation courses, and little or no reactive cell infiltrate was still evident. In only a few mammary glands, residual and constrained foci of hyperplasia associated with a distinct reactive cell infiltrate were present close to the nipple (data not shown).
Immunological Events Associated with Carcinogenesis Inhibition.
A compendium of the immunological reactivity found in the Spc from vaccinated BALB-neuT mice is displayed in Table 1 ⇓ . Two weeks after the first vaccination or electroporation course, fresh Spc displayed no detectable CTL activity against rp185neu+ targets (data not shown), whereas a low CTL activity only was displayed by Spc from p185EC-TM plasmid-vaccinated and -electroporated mice cultured for 6 days in MLTI with rp185neu+ tumor cells (Table 1) ⇓ . By contrast, after anti-CD3 and -CD28 stimulation, a significant number of both CD8+ and CD8− cells produced IFN-γ as assessed with the Miltenyi Biotec cell enrichment and detection kit. No higher CTL activity was observed by repeating these experiments after the four courses of immunization by electroporation (data not shown).
Two weeks after the first vaccination course, the titer of anti-rp185neu antibodies was significantly higher in the sera from the electroporated compared with vaccinated mice (Fig. 4A ⇓ ; P < 0.009). After both vaccination and electroporation, anti-rp185neu antibodies were mainly of the IgG subclasses, with IgG2a as the most and IgG1 as the least represented (Fig. 4B) ⇓ . Sera from electroporated mice show a significantly higher titer of IgG2a (P < 0.0001), IgG2b (P < 0.0001), and IgG3 (P < 0.003) from that observed in the vaccinated mice. Because these IgG isotypes mediate antibodydependent cellular cytotoxicity, their ability to guide Spc lytic activity was assayed in vitro. The sera from electroporated mice guided a marked antibody-dependent cellular cytotoxicity against rp185neu+ target cells (Table 1) ⇓ .
Considering that sequential courses of DNA electroporation are required to restrain the progression of the preneoplastic lesions, the kinetics of antibody titer were evaluated. Eight weeks after each electroporation course, the titer of anti-rp185neu antibodies decreased, whereas it was boosted (P < 0.0001) by the following course (Fig. 4C) ⇓ . An increase in CD4+ cells was also evident in the Spc of mice that received the four electroporation courses (Table 2) ⇓ . After a MLTI, the amount of IFN-γ released by Spc from mice that received the four electroporation courses was significantly higher than that released after one. In agreement with the immunohistochemistry findings (Fig. 3H) ⇓ , the majority (63–68%) of cells that produced IFN-γ was CD8+ lymphocytes. By contrast, no higher CTL activity was observed after the four courses of immunization by electroporation (data not shown).
DNA Electroporation Does Not Protect BALB-neuT/γKO and BALB-neuT/μKO Mice.
The production of antibodies to rp185neu and the secretion of IFN-γ were the two prominent immune activities associated with the clearance of Her-2/neu lesions. Because selective depletion of immune functions through antibody administration is awkward in experiments lasting 1 year, BALB-neuT/μKO mice (4) that do not produce immunoglobulin in response to p185EC-TM plasmid electroporation (data not shown), and BALB-neuT/γKO that do not produce IFN-γ (16) , were bred to weigh the in vivo importance of these immune mechanisms. Although the kinetics of Her-2/neu tumor onset in untreated BALB-neuT/μKO and BALB-neuT/γKO were similar to that of BALB-neuT mice, electroporation did not significantly protect either strain (Fig. 1, E and F) ⇓ . No anti-rp185neu antibodies were detectable in the sera of any BALB-neuT/μKO mice 2 weeks after the electroporation course (data not shown).
Adoptive Transfer of Protection.
Because the whole mount provides a picture of the stage of carcinogenesis, we exploited this as an intermediate end point to assess the protection afforded by the transfer of antibody and T lymphocytes from DNA-electroporated mice in otherwise untreated BALB-neuT recipients. Spleen T cells and pooled sera from three groups of 15-week-old, DNA-electroporated BALB-neuT mice were transferred to untreated BALB-neuT mice at weeks 10, 12, and 13. Four weeks after the last transfer, the stage of carcinogenic progression was evaluated through a computer-aided image analysis of the mammary gland whole mounts (Table 3) ⇓ . Transfer of either T cells or sera provided significant inhibition, and their combination produced a marked additive effect. No inhibition followed the transfer of serum and T lymphocytes from age-matched BALB-neuT mice electroporated with the empty vector.
BALB-neuT mice are genetically predestined to develop lethal invasive carcinomas in their mammary glands by week 22. The early and diffuse overexpression of rp185neu that occurs in suckling BALB-neuT mice (10) hinders the induction of a protective immunity against a challenge of rp185neu+ transplantable tumors (7) . Nevertheless, repeated DNA electroporations of p185EC-TM plasmids begun when multiple in situ carcinomas were present slowly cleared the preneoplastic lesions and maintained all 1-year-old BALB-neuT mice free from detectable disease.
As the transforming oncogene is embedded in the genome of these mice, a unique dynamic relationship between the oncogenic signals and inhibitory immune reactions is taking place. In BALB-neuT mice, i.m. DNA vaccination triggers a reaction that inhibits the progression of neoplastic lesions but is unable to elicit their regression. As immunity fades, lesions progress, and none of the 1-year-old mice is tumor free. As expected (19 , 20) , transfection of muscle cells and expression of the plasmid-encoded proteins are greatly increased by electroporation. As compared with simple i.m. injection, vaccination by electroporation provides a greater and more persistent mass of antigen available for the induction of immune responses. Transfection of other cells, such as antigen-presenting cells, may also be facilitated. Direct application of an electric field to the tissue could result in an inflammatory response that may further enhance the immunogenicity of the plasmid-coded antigen (21) .
In the absence of the continuous offsetting by neoplastic stem cells, the clearance of neoplastic lesions that follows a single DNA electroporation course would coincide with a definitive cure. To keep all 1-year-old BALB-neuT mice free of tumors, repeated courses of DNA electroporation are required. About 3 months after each course, the immune response declines, and carcinogenesis is no longer controlled.
By contrast with transplantable tumors, clearance of in situ carcinoma and inhibition of carcinogenesis are unaccompanied by marked CTL reactivity, and both the lack of protection in BALB-neuT/μKO mice and data from the adoptive experiments suggest that anti-rp185neu antibody plays a significant protective role, presumably because rp185neu is both the target tumor antigen and a receptor regulating cell growth (22) . By down-regulating rp185neu expression (23) , and impeding the formation of homo or heterodimers that transduce proliferative signals (24) , the antibody impedes neoplastic proliferation. The membrane down-modulation of rp185neu, its intracytoplasmic confinement, and the morphological features of inhibited proliferation associated with diminished nuclear positivity of proliferating cell nuclear antigen in the mammary glands of DNA-electroporated mice endorse this noncytotoxic role of the anti-rp185neu antibody, which may also activate complement-mediated lysis as suggested by their isotype. The ability of anti-rp185neu antibody to mediate antibody-dependent cellular cytotoxicity is another mechanism for cooperation between antibody and intratumor delayed type hypersensitivity (22) .
The inability of DNA electroporation to protect BALB-neuT/γKO mice points to a critical role of IFN-γ. IgG2a is by far the most predominant isotype of the anti-rp185neu antibody elicited, and T cell-released IFN-γ is the main IgG2a switch factor (25) . Moreover, clearance of in situ carcinomas is associated with a massive infiltration of IFN-γ-releasing T cells that penetrate the basement membrane and interact with tumor cells. Intratumor IFN-γ activation of many proinflammatory and antitumor cell activities leads to both tumor rejection and induction of an efficient immune memory (25) , even in the absence of a significant CTL response (26) . After anti-CD3 and -CD28 stimulation, a significant number of both CD8+ and CD8− cells produced IFN-γ. In a similar way, both CD4+ and CD8+ T cell-depleted Spc populations specifically released IFN-γ after expansion in the presence of mitomycin-C treated p185neu+ cells, as assessed in ELISPOT (data not shown). Intratumor IFN-γ also directly inhibits the proliferation of p185neu+ tumor cells and their production of proangiogenic factors (27) .
As numerous peptides potentially fitting in the H-2d glycoproteins are already present in the protein encoded by p185EC-TM plasmid (28) , plasmids coding for the full-length rp185neu were not used. This rules out the possibility that cells which take up plasmid DNA receive positive growth signals. Moreover, concerns over the use of the full-length Her-2/neuoncogene as a DNA vaccine also include the risk of induction of autoimmunity against the intracellular domain that is highly conserved by members of the epidermal growth factor receptor family (28) .
The combination between cellular and humoral reactivity improves inhibition of carcinogenesis in BALB-neuT mice receiving the adoptive transfer of sera and T lymphocytes. The collaboration of both humoral and cellular immune response in the rejection of Her-2/neutumors in transgenic mice has been clearly documented (29) . This combination was also required for the eradication of rp185neu+ tumors transplanted in wild-type BALB/c mice, for which rp185neu is a xenogeneic nontolerated model tumor antigen (14) . Although vaccination cured all wild-type BALB/c mice, no cure was observed in the absence of antibody, in FcγRI/III KO mice and both mice with a deficient CD8 lymphocyte effector function and those with IFN-γ KO gene.
DNA-electroporated BALB-neuT mice necroscopied at week 52 were free from overt signs of autoimmune lesions in the heart, kidney, and liver (data not shown), even if the induced anti-rp185neu antibodies cross-reacted with mouse p185neu. As autoimmunity cannot easily be dissociated from effective tumor immunity (30) , the absence of obvious autoimmune lesions may be attributable to the combination of the very poor expression of p185neu by the tissues of adult mice and the inability of BALB-neuT mice to generate a high-affinity immune response to the early and diffusely expressed transgenic rp185neu protein. An efficient immune response rests on high-avidity reaction mechanisms (3) . Even so, a low-avidity response may be effective in tumor rejection and discriminate between quantitative differences in the expression of the target antigen (31) . The response to rp185neu in BALB-neuT mice generated by DNA electroporation appears to successfully control the slow but devastating progression of multiple preneoplastic lesions and even induce their long-lasting regression without causing evident autoimmune aggression of the normal tissues where mouse p185neu or rp185neu is expressed at a much lower level.
A similar vaccination could be considered in the management of early lesions expressing one of the many deregulated oncogenic protein kinase membrane receptors directly involved in cell carcinogenesis (32) . These receptors are suitable targets for both T-cell and antibody immune reactivity. Moreover, their interaction with antibody may regulate tumor cell behavior in ways not shared by conventional tumor antigens with no biological role in malignancy (22) . As amplification of p185neu expression in premalignant lesions starts at the hyperplasia stage (33) , the most important advantage to be gained comes from attacking the target receptor before initiation of the series of genetic and epigenetic events that lead to genetic instability and result in invasive and metastatic malignancy.
The high efficacy of DNA electroporation (19, 20, 21) makes it an attractive regimen for extrapolation to a clinical setting. Although the efficacy of i.m. vaccination can be enhanced by the concurrence of costimulatory molecules and cytokines (9 , 34) , DNA electroporation provides a relatively simple method for inducing a strong protection. The lower amount of DNA required as compared with i.m. DNA vaccination, the positive results obtained in large animals (35) , along with the availability of devices for electroporation in humans, 7 could make this translation not too unlikely. Enhancement of the protection elicited by DNA electroporation through the concurrence of various accessory signals is being investigated.
Grant support: The Italian Association for Cancer Research, the Italian Ministries for the University and Health, the University of Torino, Compagnia di San Paolo, and the Center of Excellence on Aging, University of Chieti.
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 accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Note: E. Quaglino and M. Iezzi contributed equally to this work.
Requests for reprints: Guido Forni, Department of Clinical and Biological Sciences, Ospedale San Luigi Gonzaga, I-10043 Orbassano, Italy. Phone: 39 0116705418; Fax: 39 0119038639; E-mail:
- Received September 19, 2003.
- Revision received December 23, 2003.
- Accepted February 16, 2004.
- ©2004 American Association for Cancer Research.