Enhancing Major Histocompatibility Complex Class I Antigen Presentation by Targeting Antigen to Centrosomes¹

DCs⁴ are the most potent professional antigen presenting cells that prime helper and killer T cells in vivo (for review see Refs. (1, 2, 3). Vaccine strategies that target DCs provide an opportunity to enhance T-cell mediated immunity. Efficient delivery of DNA directly into professional antigen-presenting cells in vivo can be achieved through the use of the gene gun, which sends DNA-coated gold beads into the epidermis (4). Our lab has successfully used the gene gun to test several strategies for enhanced antigen presentation (5, 6, 7, 8, 9).

One novel strategy to enhance the MHC class I presentation of antigen is to target antigen to the centrosome. The centrosome, also called the microtubule organizing center, is a perinuclear organelle that contains a high density of proteasomes (10, 11, 12). Several other proteins, including γ-tubulin and β-tubulin, are also localized in the centrosome. The centrosome has been implicated as an important intracellular compartment for proteasomal degradation of proteins into antigenic peptides (10). We hypothesize that a strategy that targets and concentrates antigens to the centrosome may be able to facilitate proteasomal degradation of antigen and to enhance MHC class I presentation of antigen, resulting in improved vaccine potency.

We, therefore, created a novel fusion of γ-tubulin with a model antigen, HPV-16 E7, in a DNA vaccine. We chose HPV-16 E7 as a model antigen for our vaccine development because HPVs, particularly HPV-16, are associated with most cervical cancers and E7 is an oncogenic HPV protein that is important in the induction and maintenance of cellular transformation and is expressed in most HPV-containing cervical cancers. We found that the linkage of γ-tubulin to E7 was able to target and concentrate E7 to the centrosome, to enhance E7-specific CD8⁺ T-cell immune responses in a TAP-1-dependent manner, and to generate a potent CD4-independent antitumor effect in vaccinated mice.

For the generation of pcDNA3-E7, the DNA fragment encoding HPV-16 E7 was amplified by PCR using the template, pCMVneoBam-E7 (13) and a set of primers: 5′-ggggaattcatgcatggagatacaccta-3′ and 5′-ggtggatccttgagaacagatgg-3′. The amplified product was further cloned into the EcoRI and BamHI sites of pcDNA3 (Invitrogen, Carlsbad, CA). For construction of pcDNA3-γ-tubulin, the DNA fragment containing γ-tubulin was amplified with a set of primers, 5′-gggctcgagatgccgagggaaatcatcac-3′ and 5′-gatgaattcctgctcctgggtgcccca-3′, and with the pH3-16 plasmid (14) as DNA template. The pH3-16 plasmid was a gift kindly provided by Dr. Berl R. Oakley (The Ohio State University, Columbus, OH). The PCR product was further cloned into the XhoI and EcoRI sites of pcDNA3. For construction of pcDNA3-γ-tubulin/E7, the PCR product of γ-tubulin was further cloned into the XhoI and EcoRI sites of pcDNA3-E7. For the generation of pcDNA3-E7/GFP, the DNA fragment encoding GFP was first amplified with PCR using the template, pEGFPN1 DNA (Clontech, Palo Alto, CA) and a set of primers as follows: 5′-atcggatccatggtgagcaagggcgaggag-3′and 5′-gggaagctttacttgtacagctcgtccatg-3′. The amplified product was then cloned into the BamHI and HindIII sites of pcDNA3-E7. To generate pcDNA3-γ-tubulin/E7/GFP, the DNA fragment encoding γ-tubulin was amplified and cloned into the XbaI/EcoRI sites of pcDNA3-E7/GFP. For the generation of pSC11MCS2-γ-tubulin/E7, pcDNA3-tubulin/E7 was cut with XbaI/HindIII and cloned into the pGEM vector. pGEM-γ-tubulin/E7 was then cut with SphI/HindIII and cloned into the pSC11MCS2 vector. The accuracy of all of the constructs was confirmed by DNA sequencing.

Recombinant γ-tubulin/E7 vaccinia virus was generated using pSC11MCS2-γ-tubulin/E7 vector according to a previously described protocol (13). Plaque-purified recombinant vaccinia viruses were tested for the presence of HPV-16 E7 genome by PCR and for the expression of HPV-16 E7 protein by immunofluorescence staining.

Six- to 8-week-old female C57BL/6 mice were purchased from the National Cancer Institute (Frederick, MD) and stored in the oncology animal facility of the Johns Hopkins Hospital (Baltimore, MD). TAP-1 KO C57BL/6 mice were obtained from The Jackson Laboratory (Bar Harbor, ME). All of the animal procedures were performed according to approved protocols and in accordance with recommendations for the proper use and care of laboratory animals.

The production and maintenance of TC-1 cells has been described previously (15). In brief, HPV-16 E6, E7, and ras oncogene were used to transform primary C57BL/6 mice lung epithelial cells to generate TC-1.

Preparation of DNA-coated gold particles and gene gun particle-mediated DNA vaccination was performed using a helium-driven gene gun (Bio-Rad, Hercules, CA) according to a previously described protocol (6). DNA-coated gold particles (1-μg DNA/bullet) were delivered to the shaved abdominal region of C57BL/6 mice or TAP-1 KO mice (5 mice/group) using a helium-driven gene gun (Bio-Rad, Hercules, CA) with a discharge pressure of 400 p.s.i.

293 D^b,K^b cells (16), transfected with either pcDNA3-E7/GFP or pcDNA3-γ-tubulin/E7/GFP, were cultured for 24–36 h and then were cytospined to glass slides. Cells were fixed with 4% paraformaldehyde in 1× PBS for 30 min at room temperature, made permeable with 1× PBS containing 0.05% saponin and 1% BSA. Cells transfected with pcDNA3-E7/GFP were incubated with antimouse β-tubulin Cy3 conjugate monoclonal antibody (Sigma, Saint Louis, MO) at a concentration of 1 μg/ml for 30 min at room temperature. Unbound antibodies were removed by washing three times in 1× PBS. Samples were acquired with the Noran Oz confocal laser scanning microscope system using Invertension software (version 6.5). Slides were imaged using an Olympus IX-50 inverted microscope (×250).

Cell surface marker staining of CD8 and intracellular IFN-γ staining followed by FACScan analysis were performed using conditions described previously (6). The immortalized DC line was kindly provided by Dr. Kenneth Rock (University of Massachusetts, Worcester, MA; Ref. 17). With continued passage, we have generated subclones of DCs that are easily transfected using LipofectAMINE.⁵ DCs were transfected with γ-tubulin/E7, E7, or empty plasmid DNA using LipofectAMINE 2000 (Life Technologies, Inc., Rockville, MD). DCs were collected 24 h after transfection. 2 × 10⁵ transfected DCs (stimulators) were mixed with 2 × 10⁶ D^b-restricted E7-specific CD8⁺ T cells (Ref. 18; responders) for 16 h before intracellular cytokine staining and flow cytometry analysis were performed.

We also characterized the ability of γ-tubulin/E7 vaccinia-infected bone marrow-derived DCs from either wild-type or TAP-1 KO C57BL/6 mice to activate an E7-specific CD8⁺ T cell line using intracellular cytokine staining followed by flow cytometry analysis. Bone marrow-derived DCs were obtained by culturing bone marrow cells in the presence of granulocyte-macrophage colony-stimulating factor as described previously (19). 2 × 10⁵ DCs from wild-type or TAP-1 KO C57BL/6 mice were infected with wild-type vaccinia or γ-tubulin/E7 vaccinia at a multiplicity of infection of 10 overnight. An E7-specific T cell line served as responder cells. Vaccinia-infected DCs were washed twice and incubated with 2 × 10⁶ E7-specific CD8⁺ T cells for 16 h before intracellular cytokine staining and flow cytometry analysis were performed.

Cell surface marker staining of CD8 or CD4 and intracellular cytokine staining for IFN-γ and FACScan analysis were performed using conditions described previously (6). Wild-type or TAP-1 KO C57BL/6 mice (5/group) were immunized with 2 μg of various DNA vaccines and received a booster with the same regimen 1 week later. Before FACScan, 3 × 10⁵ splenocytes were isolated from vaccinated mice and incubated for 16 h with either 1 μg/ml E7 peptide (aa 49–57) containing an MHC class I epitope (20) for detecting E7-specific CD8⁺ T cell precursors or 10 μg/ml of E7 peptide (aa 30–67) containing an MHC class II epitope (21) for detecting E7-specific CD4⁺ T cell precursors. IFN-γ-secreting activated mouse splenocytes, MiCK-1 (PharMingen, San Diego, CA), were used as a positive control for CD4⁺ IFN-γ-secreting T cells.

For the tumor protection experiment, C57BL/6 mice (5/group) were vaccinated via gene gun with 2 μg of pcDNA3 (no insert), γ-tubulin, E7, or γ-tubulin/E7 DNA. One week later, mice were boosted with the same regimen as the first vaccination. One week after the last vaccination, mice were s.c. challenged with 5 × 10⁴ cells/mouse TC-1 tumor cells in the right leg and then monitored twice a week.

For the tumor treatment experiment, C57BL/6 mice (5/group) were i.v. challenged with 1 × 10⁴ cells/mouse TC-1 tumor cells via tail vein on day 0. Three days after challenge with TC-1 tumor cells, mice were given 2 μg of pcDNA3 (no insert), γ-tubulin, E7, or γ-tubulin/E7 DNA via gene gun. One week later, these mice were boosted with the same regimen as the first vaccination. Mice were sacrificed and lungs were explanted on day 21. The pulmonary nodules in each mouse were evaluated and counted by experimenters blinded to sample identity.

In vivo antibody depletions have been described previously (15). Briefly, C57BL/6 mice (5/group) were vaccinated with 2 μg of pcDNA3-γ-tubulin/E7 DNA via gene gun, boosted 1 week later with the same dose, and challenged with 5 × 10⁴ cells/mouse TC-1 tumor cells s.c. 1 week after the last vaccination. Depletions were started 1 week before tumor challenge. Monoclonal antibody GK1.5 was used for CD4 depletion, monoclonal antibody 2.43 was used for CD8 depletion, and monoclonal antibody PK136 was used for NK1.1 depletion. The completeness of depletion was checked by flow cytometry analysis. For each time point of analysis, >99% depletion of the appropriate subset was achieved with normal levels of the other subsets. Depletion was terminated on day 90 after tumor challenge.

All of the data expressed as means ± SE are representative of at least two different experiments. Data for intracellular cytokine staining with flow cytometry analysis and tumor treatment experiments were analyzed by ANOVA. Comparisons between individual data points were made using a student’s t test. Kaplan-Meier survival curves for tumor protection experiments and in vivo antibody depletion experiments were applied for differences between curves. Ps were calculated using the log-rank test. P < 0.05 was considered significant.

γ-Tubulin has been described as a centrosomal marker, whereas E7 typically shows a cytoplasmic and nuclear distribution. To determine whether a chimera of γ-tubulin and E7 accumulates at the centrosome, we linked the DNA encoding GFP to the 3′ end of E7 DNA or chimeric γ-tubulin/E7 DNA in the pcDNA3 vector. Transfection and subsequent examination with a confocal fluorescent microscope revealed that cells transfected with E7/GFP DNA showed homogeneous cytoplasmic/nuclear distribution (monoclonal antibody), whereas cells transfected with γ-tubulin/E7/GFP DNA displayed a perinuclear speckled pattern consistent with centrosome localization (Fig. 1,E). To confirm whether the γ-tubulin/E7/GFP chimera had in fact been distributed to the centrosome, we performed immunofluorescent staining of cells transfected with either wild-type E7/GFP or γ-tubulin/E7/GFP using an antibody against β-tubulin (Fig. 1, A and D), or vimentin (not shown), both well-characterized markers for the centrosome (22, 23). As shown in Fig. 1, C and F, colocalization with the β-tubulin protein was observed in cells transfected with γ-tubulin/E7/GFP but not in cells transfected with E7/GFP, indicating that the fusion of γ-tubulin to E7/GFP facilitated the targeting and concentration of the linked protein into the centrosome.

We reasoned that the ability of γ-tubulin/E7 to target antigen to the centrosome may enhance MHC class I presentation of E7 antigen. Thus, to assess the effect of γ-tubulin/E7 DNA on antigen presentation, we transfected DCs with empty plasmid, γ-tubulin, E7, or γ-tubulin/E7 DNA and incubated the cells with an H-2D^b-restricted E7-specific CD8⁺ T-cell line (18) followed by flow cytometry analysis. As shown in Fig. 2 A, DCs transfected with γ-tubulin/E7 DNA generated a significantly greater number of IFN-γ-secreting E7-specific CD8⁺ T cells than did DCs that were transfected with wild-type E7 DNA (one-way ANOVA on pooled splenocytes, P < 0.0001). Our results suggested that the linkage of γ-tubulin to E7 leads to the enhanced presentation of E7 antigen through the MHC class I pathway to CD8⁺ T cells.

We evaluated whether the observed enhancement of MHC class I presentation in vitro was indicative of augmentation of CD8⁺ T-cell immune responses in vivo. To determine whether vaccination of mice with pcDNA3-γ-tubulin/E7 DNA can enhance E7-specific CD8⁺ T-cell activity, we performed intracellular cytokine staining for IFN-γ-secreting E7-specific CD8⁺ T-cell precursors using splenocytes from vaccinated mice. Mice vaccinated with γ-tubulin/E7 DNA exhibited a marked increase in E7-specific IFN-γ⁺ CD8⁺ T-cell precursors compared with mice vaccinated with wild-type E7 DNA (Fig. 2 B). These results indicated that the linkage of γ-tubulin to E7 significantly enhanced E7-specific CD8⁺ T-cell-mediated immune responses and that fusion of E7 to γ-tubulin was essential for this observed enhancement because γ-tubulin mixed with E7 (γ-tubulin+E7 DNA) did not enhance CD8⁺ T cell activity. Furthermore, the linkage of irrelevant proteins (such as GFP) to E7 did not lead to the enhancement of E7-specific CD8⁺ T-cell activity (data not shown).

Although linkage of γ-tubulin to E7 led to enhanced E7-specific CD8⁺ T-cell activity, we did not detect a significant difference in the number of E7-specific IFN-γ-secreting CD4⁺ T cells (Fig. 2 C) among the various vaccination groups (one-way ANOVA on pooled splenocytes, P = 0.2367). When the splenocytes collected from γ-tubulin/E7-vaccinated mice were pulsed with the E7_30–67 peptide, we observed a significant population of CD4-negative IFN-γ-secreting cells by flow cytometry analysis. Because the E7_30–67 peptide also contains the E7 H-2D^b-restricted epitope (aa 49–57), the majority of the CD4 negative IFN-γ-secreting cells are likely E7-specific IFN-γ-secreting CD8 T+ cells.

To determine whether the observed enhancement in E7-specific CD8⁺ T-cell-mediated immunity translated to a significant E7-specific antitumor effect, we performed an in vivo tumor protection experiment using a previously characterized E7-expressing tumor model, TC-1 (15). As shown in Fig. 3 A, 100% of mice receiving the γ-tubulin/E7 DNA vaccine remained tumor free 63 days after the TC-1 challenge. In contrast, all of the unvaccinated mice and mice receiving pcDNA3 (no insert), γ-tubulin alone, or wild-type E7 developed tumor growth within 14 days of tumor challenge. We also observed that the fusion of E7 to γ-tubulin was required for antitumor immunity, because γ-tubulin mixed with E7 (γ-tubulin+E7 DNA) failed to enhance tumor protection compared with E7 DNA alone (log-rank test, P < 0.0001). These data suggested that the physical linkage of γ-tubulin to E7 was important for the observed enhancement of E7-specific tumor protection.

We also investigated the therapeutic potential of the chimeric γ-tubulin/E7 DNA construct in treating TC-1 tumor metastases in the lungs. As shown in Fig. 3 B, mice treated with γ-tubulin/E7 DNA exhibited the lowest mean number of pulmonary nodules (0.75 ± 0.95) compared with mice treated with wild-type E7 DNA (62.6 ± 3.5), or γ-tubulin DNA (94 ± 2.5; one-way ANOVA, P < 0.001). The results from the tumor protection and treatment experiments indicated that the linkage of γ-tubulin to E7 dramatically enhanced antitumor effects against the growth of TC-1 tumors.

We performed in vivo antibody depletion experiments to determine the subset of lymphocytes important for the antitumor effect (24). As shown in Fig. 3 C, all of the naïve mice and all of the mice depleted of CD8⁺ T cells grew tumors within 15 days of tumor challenge. In contrast, all of the nondepleted mice and the mice depleted of CD4⁺ T cells or NK1.1 cells remained tumor-free 60 days after tumor challenge (log-rank test, P < 0.0001). These results suggested that the antitumor effect generated by the γ-tubulin/E7 DNA vaccine was dependent on CD8⁺ T-cell immune responses but independent of CD4- or NK-associated responses.

To elucidate the role of TAP-1-dependent processing on the observed E7-specific CD8⁺ T-cell-mediated immune response in γ-tubulin/E7 DNA-vaccinated mice, we used TAP-1 KO C57BL/6 mice. TAP-1 KO C57BL/6 mice failed to generate appreciable E7-specific CD8⁺ T-cell immune responses (Fig. 4,A). To further evaluate the importance of TAP-1 for the presentation of E7 antigen to E7-specific CD8⁺ T cells in vaccinated mice, we isolated bone marrow-derived DCs from wild-type or TAP1 KO C57BL/6 mice and infected these cells with either wild-type vaccinia or γ-tubulin/E7 vaccinia for use as stimulators. Vaccinia-infected DCs were incubated with an E7-specific CD8⁺ T cell line. We found that γ-tubulin/E7 vaccinia-infected DCs from wild-type C57BL/6 mice were able to efficiently induce E7-specific CD8⁺ T cells to secrete IFN-γ; in contrast, γ-tubulin/E7 vaccinia-infected DCs from TAP-1 KO C57BL/6 mice were unable to induce E7-specific CD8⁺ T cells to secrete IFN-γ (Fig. 4 B). These data indicated that the observed enhancement of E7-specific CD8⁺ T-cell immune responses in γ-tubulin/E7 DNA-vaccinated mice was dependent on the TAP-1-dependent MHC class I presentation pathway.

In this study, we demonstrated that the direct linkage of γ-tubulin to E7 targets and concentrates E7 to the centrosome. This approach dramatically enhanced the potency of HPV-16 E7-containing DNA vaccines, leading to a marked enhancement of E7-specific CD8⁺ T-cell immune responses and antitumor effects.

Our results demonstrated that TAP-1 was essential for the enhancement of CD8⁺ T-cell immune responses in γ-tubulin/E7 DNA-vaccinated wild-type C57BL/6 mice because CD8⁺ T-cell immune responses were significantly weaker in γ-tubulin/E7 DNA-vaccinated TAP-1 KO C57BL/6 mice. TAP-1 and TAP-2 are integral ER proteins with an ATP-dependent transport function that shuttle oligopeptides from the cytoplasm to the ER (25, 26, 27). Our study suggests that peptides that are processed in proteasomes of the centrosome also depend on TAP-1 for association with MHC class I molecules in the ER and antigen presentation on the cell surface. Although TAP-1-independent antigen processing pathways have been described (28, 29, 30, 31), these pathways apparently do not contribute substantially to the immune response given the weak immune response observed in TAP-1 KO mice.

Our data indicate that the centrosome is a potentially important locus for the processing of antigenic peptides for MHC class I presentation. This perinuclear structure is a site for the accumulation of proteasomes and associated regulatory proteins including ubiquitin (32, 33) and HSPs-70 and -90 (11, 12) and facilitates intracellular protein degradation. Studies have demonstrated that proteasomal inhibitors hinder processing by cytosolic (34, 35) or centrosomal proteasomes (10) and encumber the stable assembly of MHC class I molecule/peptide complexes. Another study found that proteasomal degradation of misfolded antigen (denatured influenza nucleoprotein), and, presumably, generation of antigenic peptides, occurs in the centrosome and promyelocytic leukemia oncogenic domains (10). It is likely that proteasomes in the centrosome catalyzed the degradation of γ-tubulin/E7 protein and the generation of epitopes for MHC class I presentation. Because of its abundance of proteasomes and proximity to the ER, the centrosome may be a key site for the processing of certain antigens and subsequent presentation through the MHC class I pathway.

The success of the γ-tubulin/E7 DNA vaccine warrants the consideration of strategies that use other proteins that also target the centrosome to enhance vaccine potency. Biochemical analysis of the centrosome has revealed that a variety of proteins including β-tubulin (22), pericentrin (36), HSP-70 (11), HSP-90 (11), human TOGp (37, 38), p115 (39), and the homologues of these proteins are targeted and concentrated to the centrosome. Understanding the targeting nature of these proteins would facilitate their application in vaccine developments in a manner similar to that described in the present study.

We have used gene gun immunization to test several intracellular targeting strategies that enhance MHC class I and/or class II presentation of antigen. In addition to the γ-tubulin/E7 DNA vaccine strategy, MHC class I presentation of HPV-16 E7 can be significantly enhanced by linkage with Mycobacterium tuberculosis HSP (HSP-70; Ref. 6) or the translocation domain (domain II) of Pseudomonas aeruginosa exotoxin A [ETA(dII); Ref. 8] in the context of a DNA vaccine. A comparison of γ-tubulin/E7 DNA, E7/HSP70 DNA, and ETA(dII)/E7 DNA revealed that the enhancement of E7-specific CD8+ T-cell immune responses was slightly greater in mice vaccinated with γ-tubulin/E7 DNA compared with the other vaccinated mice and wild-type control (data not shown). Whereas chimeric γ-tubulin/E7, E7/HSP70, ETA(dII)/E7, or CRT/E7 DNA generates potent CD8+ T-cell responses through enhanced MHC class I presentation, other constructs that target antigen to MHC class II presentation pathways may provide enhanced CD4⁺ T-cell responses. For example, a DNA vaccine encoding a signal sequence linked to E7 and the sorting signal of the lysosome-associated membrane protein (LAMP-1; creating the Sig/E7/LAMP-1 chimera) targets E7 to endosomal and lysosomal compartments and enhances MHC class II presentation to CD4⁺ T cells compared with DNA encoding wild-type E7 (5). The availability of these strategies raises the notion of coadministration of vaccines such as of γ-tubulin/E7 DNA and Sig/E7/LAMP-1 DNA in a synergistic fashion. Such an approach may directly enhance both MHC class I and class II presentation of E7 and may lead to significantly enhanced E7-specific CD4⁺ and CD8⁺ T-cell responses and potent antitumor effects.

In summary, our findings indicate that the linkage of γ-tubulin to antigen is able to target antigen to the centrosome and elicit potent antigen-specific CD8⁺ T-cell activity and antitumor effects in vivo. This approach may be promising for future vaccine development to control other cancers or infectious diseases and may be particularly useful in patients with diminished CD4⁺ T-cell immune responses, particularly HIV-infected patients.

We thank Drs. Robert J. Kurman, Ken-Yu Lin, and Drew Pardoll for insightful discussions. We also thank Dr. Richard Roden for critical review of the manuscript.