| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of
*
Dermatology, *Immunology, and
Molecular Genetics and Biochemistry, University of Pittsburgh School of Medicine, and University of Pittsburgh Cancer Institute, Pittsburgh, PA 15261
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
by Ag-specific T cells. In the B16-OVA tumor therapy model, the growth of established tumors was significantly inhibited by a single immunization using lentiviral vector-transduced DCs, resulting in significantly longer survival of immunized animals. These results suggest that compared with Ag-pulsed DCs, vaccination with lentiviral vector-transduced DCs may achieve more potent antitumor immunity. These data support the further development of lentiviral vectors to transduce DCs with genes encoding Ags or immunomodulatory adjuvants to generate and control systemic immune responses. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
100 clinical trials aimed at inducing antitumor immunity, the majority of which use DCs loaded with peptide or protein tumor Ags (11, 12, 13, 14, 15). Several studies suggest that surface presentation of MHC class I peptide complexes by peptide- or protein-loaded DCs is relatively short-lived (16, 17, 18, 19), potentially limiting the potency of the resulting immune responses. In addition, recent studies raise the possibility that Ag from Ag-loaded DCs can be cross-presented by resident DCs, contributing to the generation of the immune response (20, 21, 22). Notably, apoptosis of DCs expressing transgenic Ags has been shown to result in enhanced immunity through cross-presentation in vivo (23). Furthermore, the latest studies suggest that cross-presentation favors epitopes derived from stabilized proteins rather than those from rapidly degraded proteins or short peptides (24, 25, 26). Given these observations, it is reasonable to suggest that the use of DCs presenting a sustained source of endogenously synthesized transgenic protein Ag may be advantageous for T cell induction due to both the prolonged presentation by modified DCs and cross-presentation by endogenous DCs.
Viral transduction is an appealing approach for engineering DCs to produce, process, and present specific Ags (27). Adenoviral vectors have been shown to be effective means for transducing DCs (28, 29, 30). However, recent data indicate that adenoviral transduction results in DC maturation and proinflammatory cytokine production through activation of the NF-
B pathway (31). In addition, Ags derived from adenoviral proteins can be presented by transduced DCs, resulting in not only the interference of the desired Ag-specific immune responses, but also the induction of vector-specific immunity that makes repeated immunization problematic (32). Oncoretroviral vectors have been used to deliver genes into DCs with minimal interference of their function, but low transduction efficiencies, presumably due to the fact that oncoretroviral vectors are unable to transduce nondividing cells, limit their applications (33).
To overcome these disadvantages, lentiviral vectors (lvv) have recently been developed to introduce foreign genes into DCs (34, 35, 36, 37, 38). In contrast to oncoretroviral vectors, lvv are able to transduce nondividing cells, such as DCs, at high transduction efficiencies. Importantly, like oncoretroviral vectors, lvv do not encode viral proteins, thereby minimizing the potential for interfering with the function of transduced DCs (39). Recently, third-generation lvv with enhanced safety profiles have been developed and used to efficiently transduce murine DCs and human DCs (36, 37, 40). These improved vectors contain a chimeric Rous sarcoma virus (RSV)/HIV 5' long terminal repeat (5'LTR) enhancer and promoter to initiate the transcription of genomic viral RNA (41). The stronger chimeric promoter does not require HIV Tat protein, a transactivator in the transcription of HIV genomic RNA, to generate vector transcripts. In addition, the vectors have been made self-inactivating by deleting the majority of U3 region in 3'LTR so that viral RNA cannot be produced in target cells (42). These additional safety modifications further prevent the generation of replication-competent recombinants. In aggregate, these features suggest that this lvv has considerable potential for clinical applications.
Although, in theory, presentation of endogenously expressed Ag by gene-modified DCs should elicit more potent immune responses than peptide- or protein-pulsed DCs, results from comparative investigations are conflicting (32, 43, 44, 45). This may be a critical issue for immunotherapy, particularly in the setting of tumor-bearing hosts or victims of viral infections such as HIV, in whom immunizations will need to overcome pre-existing immunosuppressive mechanisms. We recently demonstrated, by direct comparison, that DCs expressing transgenic Ag after nonviral transfection induced more potent T cell immunity than peptide/protein-pulsed DCs (46). As yet there are no reports comparing the immunogenicity of lvv-transduced DCs with that of peptide/protein-pulsed DCs. In this study, using a third-generation lvv optimized for safety and efficiency, we directly compare these Ag-loading strategies by comparing the kinetics and strength of Ag-specific immune responses induced in vitro and in vivo. We show that DCs transduced with lvv stimulate stronger and long-lasting in vivo CTL activity compared with peptide/protein-pulsed DCs. Furthermore, we demonstrate that DCs modified by lvv can achieve significantly better therapeutic antitumor effects compared with peptide/protein-pulsed DCs. These findings suggest that third-generation lvv have considerable potential for clinical application.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
C57BL/6, BALB/c, and OVA TCR transgenic mice (OT-I and OT-II) were purchased from The Jackson Laboratory. The mice were housed under specific pathogen-free conditions in the central animal facility of University of Pittsburgh. Handling of the mice was performed according to institutional guidelines of University of Pittsburgh.
DC generation
Mouse DCs were generated using standard protocols (47). Briefly, bone marrow cells were depleted of T, B, and NK cells with Ab and complement. Cells were then seeded at 15–20 million/75-cm flask in RPMI 1640 medium containing GM-CSF (1000 U/ml) and IL-4 (1000 U/ml). Half the medium was changed every other day. By day 8, 90% of cells were CD11c+.
Lentiviral vector production and concentration
The third-generation lvv pLenti6/-TOPO plasmid was purchased from Invitrogen-Life Technologies. Genes encoding enhanced GFP (EGFP) or truncated cytoplasmic chicken OVA (first 138 aa were deleted) were cloned into the vector downstream of the CMV promoter using BamHI and XhoI sites. Our previous studies demonstrated that this cytosolically expressed transgenic Ag can be efficiently presented (46). To enhance nuclear import, a DNA fragment containing a central polypurine tract sequence and the central termination sequence, which together form a triple-stranded DNA flap (TRIP), was inserted in the unique site of ClaI in front of CMV promoter as previously described (48). The final plasmid constructs, pLenti-EGFP-TRIP and pLenti-OVA-TRIP, were used to generate lentivirus, as shown diagrammatically in Fig. 1.
|
Transduction with lvv and Ag loading
Six million DCs or their progenitors were washed with serum-free RPMI 1640 medium once and resuspended in 200 µl of serum-free medium. Cationic polymer polybrene was added to serum-free medium to a final concentration of 32 µg/ml. One hundred microliters of lvv (1.2–2.4 x 108 TU) in PBS was mixed with 100 µl of polybrene-containing, serum-free medium and incubated at room temperature for 15 min. The cells were mixed with virus and incubate at 37°C for 3 h. The final 400 µl of transduction mixture contained 6 million DCs with 1.2–2.4 x 108 TU of lentivirus. Six milliliters of GM/IL-4-containing medium was added to each tube and divided into two wells (six-well plate). The cells were cultured at 37°C. Half the medium was replaced with fresh GM/IL-4 medium every other day. To load DCs with protein or peptide Ags, DCs were pulsed with OVA protein (200 µg/ml) and or/the class I-restricted peptide SIINFEKL (10 µg/ml) or the class II-restricted peptide ISQAVHAAHAEINEAGR peptide (10 µg/ml), as indicated, in RPMI 1640 medium for 3 h at 37°C. DCs were then washed three times before use.
MLR assay
BALB/c splenic T cells were enriched using nylon wool columns, and B cells, DC, and macrophages were depleted by anti-B220, anti-MHC II (M5114), and anti-macrophage (F4/80) Ab (American Type Culture Collection) and complement (Cedarlane Laboratories). DCs were serially diluted in a round-bottom, 96-well plate started at 2000 cells/well. One hundred microliters of BALB/c cells (0.1 million) were added to each well. Cells were cocultured for 2 days, then 100 µl of medium was collected for measuring cytokine production. [3H]Thymidine was added for an additional 24 h of culture, then cells were harvested, and cell proliferation was measured by [3H]thymidine incorporation.
OT-I and OT-II cell proliferation assay
OT-I and OT-II T cells were isolated from the spleen and lymph nodes (LN) of OVA TCR transgenic mice. The splenocytes and LN cells were enriched on nylon wool columns, then B cells, NK cells, DCs, macrophages, and CD4 or CD8 T cells were deleted using Ab and complement. The purity of OT-I and OT-II T cells was determined to be >95%. DCs (transduced or pulsed) were then serially diluted as in the MLR assay. OT-I or IT-II T cells were added to each well in 100 µl (20,000 cells/well). Cells were cocultured for 2 days, 100 µl of medium was collected for measuring cytokine production. [3H]thymidine was added for an additional 24 h of culture, then cells were harvested, and cell proliferation was measured by [3H]thymidine incorporation.
Cytokine detection
The quantity of cytokines, IL-12p70 and IFN-, were measured using the OPTIA kit from BD Pharmingen according to the manufacturer’s instructions.
Flow cytometric analysis
Cells were collected and washed with PBS. A half-million cells were added to each staining tube. Anti-CD11c-biotin, anti-B7.2-PE, anti-IFN--allophycocyanin, anti-V
2-PE, and anti-V
5.1–5.2-biotin Abs were purchased from BD Pharmingen. CyChrome- or allophycocyanin-labeled streptavidin was used as secondary reagent at a 1/1,000 dilution. Cells were fixed, and 30,000 events were collected and analyzed.
Tumor growth and vaccination
Mice were inoculated with 0.2 million B16-OVA cells in 50 µl of PBS s.c. Three days later, 500,000 lvv-transduced or SIINFEKL/OVA-pulsed DCs were injected into the footpads of the mice. Tumor growth was monitored by measuring the perpendicular diameter of the tumor, and survival was recorded daily.
In vivo killing assay
In vivo Ag-specific lytic activity was measured by an in vivo killing assay, as previously described (51). Briefly, mouse splenocytes were collected and pulsed with SIINFEKL (250 ng/ml) and were labeled using a high concentration of CFSE (5 µM; CFSEhigh). Mouse splenocytes without SIINFEKL peptide were labeled using a low concentration of CFSE (0.5 µM; CFSElow) as an internal control. Ten million cells of each population were mixed and injected into mice via the tail vein. The relative abundance of CFSEhigh and CFSElow cells in spleen or peripheral blood was determined by flow cytometry at 5 and/or 20 h after injection, as specified in the individual experiments. Specific lysis was calculated according to the following formula: {1 – [(ratio of CFSElow/CFSEhigh of naive mouse)]/[(ratio of CFSElow/CFSEhigh of vaccinated mouse)]} x 100.
Intracellular staining of IFN-
Splenocytes were collected and stimulated in vitro for 20 h by adding OVA class I peptide SIINFEKL or class II peptide ISQAVHAAHAEINEAGR. GolgiStop (BD Pharmingen) was added to the culture in the last 4 h of restimulation to block the secretion of IFN-. Cells were then collected and stained for CD8 or CD4, and intracellular IFN- was detected using an intracellular staining kit according to the manufacturer (BD Pharmingen).
In vivo Ag presentation assay
The in vivo Ag presentation assay was performed as previously described (52). Briefly, mice were injected in the footpad with 500,000 DCs that were transduced with lvv expressing OVA or pulsed with SIINFEKL peptide or OVA protein as indicated. At the indicated time points, 5 million OT-I cells labeled with CFSE were administered through the tail vein. Forty-eight hours after injection of OT-I cells, draining popliteal LN and contralateral LN cells were collected and stained with anti-V2 and anti-V5.1–5.2 and analyzed for proliferation as described previously (52).
Statistical analysis
Unpaired t test analysis was used to determine whether the differences between T cell-mediated immune responses induced by peptide/protein-pulsed DCs vs transduced DCs was significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Day 2 DCs were transduced in vitro with EGFP-lvv at 40 TU/cell On day 8, DCs were collected and stained for expression of CD11c and B7.2. Approximately 84% of the total cells in culture in both mock- and lvv-transduced cells were CD11c+, and 50% of these expressed EGFP (Fig. 2A). In addition, when the levels of expression of B7.2 were used as a marker of DC maturation, it was found that approximately equal numbers of mature (CD11c+B7.2high) and immature (CD11c+B7.2low or neg) DCs expressed EGFP (Fig. 2B). Importantly, DC transduction did not alter the relative proportions of mature vs immature DCs in DC populations (Fig. 2B), suggesting that lentiviral transduction of DC progenitors or DCs at early stages does not alter DC development or maturation.
|
90% of the cells expressed CD11c (data not shown).
|
Ag presentation function of lvv-transduced DCs
To determine whether lvv transduction affected the Ag presentation function of transduced DCs, we evaluated and compared the capacities of transduced and mock-transduced DCs to stimulate allogenic T cells. Bone marrow cell cultures were transduced on day 2 with OVA-lvv vector encoding a cytoplasmic form of OVA protein or were mock-transduced with PBS. On day 8, DCs were cocultured with BALB/c T cells at varying DC:T cell ratios. The lvv-transduced DCs induced strong allogenic T cell proliferation and IFN- secretion (Fig. 3B), comparable to those induced by mock-transduced DCs.
To investigate the capacity of transduced DCs to present transgenic Ags, DCs transduced with OVA-lvv or PBS mock transduced DCs were used to stimulate class I restricted OT-I or class II-restricted OT-II T cells isolated from OVA TCR transgenic mice. As a positive control, presentations by otherwise identical DC populations pulsed with the corresponding class I-restricted (SIINFEKL) or class II-restricted (ISQAVHAAHAEINEAGR) peptide were compared. As shown in Fig. 4, DCs transduced with OVA-lvv vector presented OVA to both class I- and class II-restricted T cells, as assessed by T cell proliferation and IFN- secretion. Remarkably, class I-restricted presentation of transgenic Ag by transduced DCs was comparable to that observed by DCs pulsed with saturating amounts of exogenous peptide (Fig. 4, A and B). Although transduced DCs presented synthesized transgenic Ag to class II-restricted T cells, presentation was less efficient than that observed by peptide-pulsed DCs (Fig. 4C; using unpaired t test, at 1:10 and 1:20, p < 0.05), presumably due to decreased access of endogenously synthesized peptide to the class II-restricted processing pathway. It is important to note that studies from other laboratories have indicated that inclusion of leader sequences targeting expression to the endoplasmic reticulum or endosome can increase the efficiency of presentation of endogenously expressed class II-restricted epitopes (53).
|
To compare the strength of T cell responses in vivo, mice were vaccinated with DCs pulsed with SIINFEKL and OVA protein (SIINFEKL/OVA-pulsed DCs) or DCs transduced with OVA-lvv. Mock-transduced DCs were used as a control. Eight days after immunization, we measured in vivo Ag-specific killing and determined intracellular IFN- expression by Ag-specific T cells. Determination of Ag-specific lytic activity from measurement of target lysis in peripheral blood 5 h after target injection demonstrated a quantitatively more potent response in OVA-lvv DC-immunized animals, with OVA-lvv-immunized animals demonstrating 4 times more Ag-specific in vivo lytic activity than mice immunized with SIINFEKL/OVA-pulsed DCs (82 vs 20% lysis, respectively; Fig. 5, A and B). Not surprisingly, both OVA-lvv-pulsed DC- and SIINFEKL/OVA-pulsed DC-immunized mice demonstrated considerable lytic activity when evaluating target cells localizing in spleen 20 h after target cell injection (99 and 91% lysis, respectively). The CTL-mediated immunity induced by peptide/protein-pulsed DCs we observed under these conditions was consistent with previously reported results of CTL assays using splenic effectors (3).
|
production, splenocytes from immunized animals were restimulated in vitro with SIINFEKL (class I) or ISQAVHAAHAEINEAGR (class II) peptide for 20 h, then stained for the expression of CD8 or CD4 and cytoplasmic IFN-. Consistent with results from the in vivo lysis assay, as shown in Fig. 5, C and D, 2% of the spleen CD8+ T cells from mice vaccinated with OVA-lvv DCs produced IFN-, whereas only 0.5% of CD8+ T cells from mice vaccinated with SIINFEKL/OVA DCs were IFN- positive. Notably, neither immunized group demonstrated significant IFN- production from the CD4+ T cell population, with <0.1% of CD4+ T cells producing IFN-. OVA-lvv-transduced DCs induce long-lasting immune responses in vivo
We compared the kinetics of T cells responses induced by OVA-lvv-transduced DCs with responses induced by DCs pulsed with either SIINFEKL peptide or OVA protein. The in vivo CTL activity induced by DCs pulsed with SIINFEKL alone peaked on day 5 and declined to undetectable levels in 2 wk (Fig. 6). Not surprisingly, OVA protein-pulsed DCs induced longer-lasting in vivo CTL activity compared with SIINFEKL-pulsed DCs. However, the in vivo CTL activity induced by OVA protein-pulsed DCs was also dramatically reduced by day 15 (Fig. 6). The magnitude and kinetics of the response induced by DCs pulsed with both peptide and OVA protein were similar to those induced by DCs pulsed with OVA protein alone (data not shown). In contrast, the mice injected with OVA-lvv DCs maintained a high level of CTL activity through day 15 and maintained significant Ag-specific lytic activity through day 41, the latest time point tested (Fig. 6). These data indicate that immunization with OVA-lvv-transduced DCs stimulated considerably longer-lasting T cell immunity than that induced by protein- or peptide-pulsed DCs.
|
One possible mechanism for the long-lasting CTL activity induced by OVA-lvv-transduced DCs is extended Ag presentation in vivo. To investigate the duration of Ag presentation in vivo, immunized mice were injected with naive OVA-specific OT-I T cells 5, 8, or 12 days after immunization. The CFSE-labeled T cells were collected from the LNs, and in vivo proliferation was determined by flow cytometry as previously described. Interestingly, through day 8 after immunization, all DCs induced similar levels of proliferation (Fig. 7). Furthermore, although OVA-specific T cell proliferation could still be detected 12 days after immunization in all groups, proliferation in mice immunized with OVA-lvv-DCs was considerably lower than that seen in mice immunized with protein- or peptide-pulsed DCs (Fig. 7).
|
To evaluate the potential use of lentiviral-transduced DCs for tumor immunotherapy in vivo, we used a well-defined and very aggressive melanoma tumor model. In this model, the B16 melanoma expresses the cytosolic form of the model Ag OVA. Similar to MO4 (5), the expression of OVA in B16-F10 cells does not increase the immunogenicity of this tumor (data not shown). Initial experiments showed that naive mice vaccinated with OVA-lvv DCs before tumor challenge were completely protected from B16-OVA tumor cell challenge and remained tumor free for at least 2 mo, unlike mice immunized with control mock-transduced DCs or EGFP-lvv DCs, which developed progressively growing tumors comparable to those in unimmunized animals (data not shown). To evaluate the therapeutic potential of this approach, we established s.c. B16-OVA tumors by inoculating 0.2 million cells intradermally into C57/BL6 mice. After 3 days, when the tumor was established with a size of at least 3–4 mm in diameter, mice were immunized by intradermal injection of either 0.5 million SIINFEKL/OVA-pulsed or OVA-lvv-transduced DCs at distant sites. As shown in Fig. 8, animals treated with SIINFEKL/OVA-pulsed DCs showed marginal inhibition of tumor growth and had survival rates no better than those of control mice. In contrast, in mice immunized with transduced DCs, tumor growth was strongly inhibited (p < 0.01 compared with control mice or mice vaccinated with pulsed DCs), and the mice survived significantly longer. In this aggressive tumor model, mice immunized with lvv-transduced DCs demonstrated a significant therapeutic benefit over those immunized with peptide/protein-pulsed DCs.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Gene delivery into DCs remains problematic. First, the fact that DCs are terminally differentiated poses a considerable obstacle for gene delivery by most vectors. Second, the function and phenotype of DCs are extremely plastic and can be readily influenced by environmental factors in the medium or by gene delivery vectors (54, 55, 56). Thus, the transfection or transduction process can introduce unwanted functions into gene-modified DCs that can limit their subsequent applications. Vectors or gene delivery strategies that achieve high efficiency without perturbing the intrinsic properties and function of DCs will contribute significantly to DC-based immunizations and immunotherapies.
Retroviral vectors, especially HIV-based lvv, may fulfill the requirements for transducing DCs (57). Lentiviral vector has the advantage of transducing nondividing cells and thus should be more effective for transducing DCs (49). Like oncoretroviral vectors, only minimal cis elements are needed for lvv production. The few essential proteins are provided through a trans-complementation mechanism encoded by separate plasmids during vector production. Thus, there are no genes encoding viral proteins in the lvv particles. The low amount of regulatory viral proteins present in the vector particles in the absence of a sustained source from transduced cells is likely to minimize vector effects on the functions of transduced DCs, providing an important advantage to lvv for introducing Ag-encoding genes into DCs (54). In the current study our data confirmed that lvv transduction of DC progenitors or immature DCs did not affect their maturation or intrinsic function (Figs. 2 and 3). Although important for immunization strategies, this may be particularly relevant for efforts to induce Ag-specific regulatory T cells, because the use of lentivectors appears to avoid the vector-mediated immunostimulatory effects of other viral vectors, such as adenovirus (31).
We directly addressed whether maturation would affect the efficiency of lvv transduction of murine DCs. Our study indicated that transduction of day 2 DC progenitors or immature DCs resulted in the best transduction efficiency (Fig. 2C). Progenitor cells cultured in the presence of GM-CSF and IL-4 appeared to be better targets than those directly isolated from bone marrow, possibly because day 2 bone marrow cells may be in a stage of proliferation secondary to stimulation by GM-CSF and IL-4 cytokines, whereas the day 0 bone marrow cells may be quiescent. When DC populations become more mature with increased time in culture, the transduction efficiency begins to decline, suggesting that mature DCs were less likely to be transduced. These data suggested that although lvv with nuclear localization sequences (TRIP) can transduce quiescent cells, dividing or proliferating cells may offer additional advantages for improving transduction efficiency. It is known that dissolution of the nuclear membrane barrier during cell division facilitates the entry of exogenous DNA into the nucleus and thus increases transduction efficacy. In addition, DNA replication assists the integration of foreign DNA into the host genome. These factors may help explain why cells in the early stage of proliferation are more readily transduced. These results from our analysis of mouse DCs resemble those of previous studies focused on human CD34+-derived DCs (58), indicating that even though lentivirus could be imported into the nucleus through its unique nuclear import protein, cell proliferation still facilitated gene delivery and expression in DCs. It is important to note that others have reported that factors such as the use of 1) higher multiplicity of infection for transduction, 2) increased concentration of vector, 3) cocentrifugation of DCs and viral particles, and 4) repeat transduction can all increase DC transduction efficiency (36, 59, 60). This suggests that the potency of the transduced DCs used in our studies could be even further enhanced by optimizing DC transduction.
Previous studies comparing the immunogenicity of peptide/protein-loaded DCs with transduced DCs have produced mixed results. Using -galactosidase as a model Ag, it has been shown that retroviral vector-transduced DCs induce stronger CTL activity and greater IFN- secretion by responding T cells compared with pulsed DCs (44). However, in the same studies there was no significant difference in antitumor activity between pulsed and transduced DCs under the conditions tested using the colon cancer CT26 lung metastasis tumor model. Using a tumor protection model, it has been shown that immunization with retroviral vector-transduced DCs can better protect mice from B16-OVA tumor challenge than immunization with peptide-pulsed DCs (45). However, tumor therapy was not evaluated. Adenoviral vector-transduced DCs have been compared with peptide/protein-pulsed DCs in two studies. Using a DC cell line, Brossart et al. (43) showed that DCs transfected with adenoviral vector or vaccinia virus expressing the SIINFEKL peptide induce stronger in vitro stimulation of CTL activity than peptide-pulsed DCs. In contrast, Tuettenberg et al. (32) recently showed that adenoviral vector-transduced DCs only induced short-lived immune responses compared with Ag-pulsed DCs in vitro. The evidence suggests that the generation of vector-specific immune responses through repeated stimulation by adenoviral vector-transduced DCs may interfere with the development and potency of the desired Ag-specific T cell responses (32). Nonviral transfection of DCs would obviate this obstacle, and the delivery of Ag-encoding mRNA to DCs has been shown to result in Ag expression and stimulation of Ag-specific T cell responses (61, 62). However, in a direct comparison, mRNA-transfected DCs were less immunogenic in vivo than lvv-transduced DCs (63). Interestingly, although introduction of mRNA into DCs by electroporation resulted in efficient transgene expression, electroporated DCs showed a 10-fold reduction in IL-12p70 secretion compared with unmodified or lentivirally transduced DCs (63). Overcoming this defect by retrovirus-mediated cotransfection of IL-12p70 overcame this limitation, but added an additional level of complexity (63).
We have recently shown that nonviral transfection of DCs can be achieved using naked DNA with very high transfection efficiencies (46). Although the use of plasmid DNA to transfect DCs results in nonintegrating transfection, we demonstrated more sustained expression of H-2Kb-SIINFEKL complexes on the surface of transfected DCs compared with DCs pulsed with SIINFEKL. In direct comparisons among peptide-pulsed DCs, naked DNA transfected DCs, and adenovirus-transduced DCs, DCs transfected with naked DNA induced considerably more potent CD8+ T cell responses in vitro and in vivo than pulsed DCs, resulting in stimulation comparable to that achieved using adenovirus-transduced DCs.
Our current studies demonstrated that in vivo, lentivector-transduced DCs can induce stronger, longer lasting T cell responses than peptide/protein-pulsed DCs. Lentiviral vector-mediated transduction results in integration of transgenes into host DNA. Although this raises concerns of potential mutagenesis, the resulting stable expression of transgenic protein may be an advantage for Ag delivery or the expression of immunomodulatory molecules in DCs to more specifically engineer immune responses.
In the current studies we compared the efficacy of T cell immune responses elicited by lentivector-transduced DCs and Ag-pulsed DCs. We demonstrate that even though in vitro, peptide- or protein-pulsed DCs and transduced DCs stimulated T cell proliferation with similar potencies, in vivo transduced DCs can induce stronger and longer lasting Ag-specific lytic T cell responses. The mechanism for this remains unclear. However, the stronger and longer lasting T cell responses we observed with lentivirus-transduced DCs are unlikely to be due an initially increased ligand density, because the MHC-SIINFEKL complexes detected by 25D1.16 Ab were not higher on OVA-lvv-transduced DCs (data not shown). It is possible that adoptively transferred DCs endogenously expressing transgenic Ag result in more durable presentation of Ag, either by longer lasting direct presentation of MHC-Ag ligand on the cell surface made possible by sustained production of Ag or by providing a more efficient source of Ag for cross-priming. Several studies suggest that diminishing class I peptide ligand density may be a limiting factor for in vivo T cell stimulation (16, 17, 18). Recently published data demonstrate that on mature DCs, five of six peptide class I ligands analyzed had cell surface half-lives of only 8.7–26.6 h when DCs were loaded by peptide pulsing (18). Additional studies suggested that although reduced membrane turnover can increase ligand longevity, newly synthesized MHC molecules do not present pulsed peptides, and it is unlikely that internal pools of pulsed peptide can contribute to ligand density (18). In the case of lvv-transduced DCs, it is likely that the cells stably express the OVA Ag, providing a sustained source of Ag to replace cycling or lost class I peptide complexes, and thus are able to generate stronger and long-lasting T cell responses. Alternatively, the more potent immune responses observed using transduced donor DCs may be a result of more efficient cross-priming by recipient DCs resident in draining lymph nodes or spleen. Several studies now suggest that cross-presentation of Ags delivered by virally infected donor DCs to recipient host lymphoid organ-resident DCs may contribute significantly to the induction of CD8+ T cell responses (20, 21, 23). Furthermore, very recent studies suggest that cross-presentation favors epitopes derived from stabilized proteins, rather than those from rapidly degraded proteins or short peptides (24, 25, 26). In this context, whole transgenic Ags produced by transduced donor DCs may serve as a more efficient source of Ag for cross-presentation than peptide Ags bound to MHC complexes or peptides or proteins internalized through the endosomal pathway of pulsed donor DCs.
Of note, using the OT-1 system to measure Ag-specific T cell proliferation, we were unable to demonstrate longer lasting in vivo Ag presentation. This system is limited however, because it is a measure of activation of naive T cells. It is possible that Ag ligand persistence sufficient to drive or help previously stimulated T cells may be important to support the long-lasting effector responses we observe, but insufficient to drive induction of the naive OT-1 T cells used in the assay. Other recent studies have shown that for maximal CTL induction, a short period of stimulation (24 h), not prolonged Ag presentation, is required (52, 64). These studies suggest that the magnitude and duration of CD8 T cell responses are determined by the initial clonal burst of T cell activation resulting from an initial period of engagement with APCs. Once the T cell immune response is initiated, the T cells are refractory to further stimulation by prolonged Ag presentation (64). Thus, it is possible that lvv-transduced DCs may enable stronger initial activation that results in stronger effector function and greater longevity. It is also possible that stronger lytic activity may eliminate some APCs more quickly (65), so that the in vivo Ag presentation measured by the naive OT-I T cell assay actually appears to be reduced. Additional experiments are needed to address these possibilities.
Importantly, our studies demonstrate that lentivirus-transduced DCs demonstrate some therapeutic efficacy in an aggressive B16 tumor model resistant to immunotherapy. B16 melanoma is a frequently studied tumor model that has been shown to escape or inhibit immune responses through a variety of mechanisms. B16 has well-document Ag escape mechanisms, including down-regulation of class I molecule expression and Ag-processing machinery (66), Growing B16 tumors have been shown to induce CD4+CD25+ regulatory cells that facilitate immune escape (67, 68, 69). In addition, B16 cells produce vesicular endothelial growth factor, which inhibits DC function and T cell immunity, and galectin-1, which is a negative regulator of T cell activation and survival (70). Taken together, these observations suggest that the B16 tumor possesses many of the tumor escape mechanisms used by human tumors and is a formidable model tumor for the evaluation of immunotherapeutic strategies. Interestingly, very recent studies suggest that sustained DC activation may overcome some forms of tumor-induced immunosuppression (71). Together with those results, our studies raise the possibility that sustained Ag expression by lentivirus-transduced DCs in combination with sustained adjuvant signaling may result in more effective immunotherapies. Lentivector-mediated expression of both Ag and costimulatory molecules or cytokines by transduced DCs could enhance or direct T cell immunity. Importantly, because transduction with lentivectors does not induce DC maturation or activation, these vectors may also provide a means to engineer DCs to induce Ag-specific tolerance through the coexpression of Ags and immunoregulatory molecules.
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
1 This work was supported by National Institutes of Health grants from the National Institute of Arthritis and Musculoskeletal and Skin Diseases (to Y.H. and L.D.F.) and the National Institute of Allergy and Infectious Diseases (to L.D.F.). 
2 Address correspondence and reprint requests to Dr. Louis D. Falo, Jr. and Dr. Yukai He, Department of Dermatology, University of Pittsburgh School of Medicine and University of Pittsburgh Cancer Institute, 190 Lothrop Street, Suite 145 Lothrop Hall, Pittsburgh, PA 15261. E-mail addresses: lof2{at}pitt.edu and E-mail address: ykhe{at}pitt.edu
3 Abbreviations used in this paper: DC, dendritic cell; EGFP, enhanced GFP; LN, lymph node; LTR, long terminal repeat; lvv, lentiviral vector; RSV, Rous sarcoma virus; TRIP, triple-stranded DNA flap; TU, transduction unit.
Received for publication July 20, 2004. Accepted for publication January 3, 2005.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
+ dendritic cells but not by Langerhans cells. Science 301:1925.B-dependent pathway. J. Virol. 74:9617.-facilitated rejection of murine B16 melanomas. J. Immunol. 161:897.-induced, CD8+ T-cell-dependent immune defense of B16 melanoma. Cancer Res. 61:8643.This article has been cited by other articles:
|
|
 |
|
  Y. Liu, Y. Peng, M. Mi, J. Guevara-Patino, D. H. Munn, N. Fu, and Y. He Lentivector Immunization Stimulates Potent CD8 T Cell Responses against Melanoma Self-Antigen Tyrosinase-Related Protein 1 and Generates Antitumor Immunity in Mice J. Immunol., May 15, 2009; 182(10): 5960 - 5969. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Loisel-Meyer, T. Felizardo, J. Mariotti, M. E. Mossoba, J. E. Foley, R. Kammerer, N. Mizue, R. Keefe, J. A. McCart, W. Zimmermann, et al. Potent induction of B- and T-cell immunity against human carcinoembryonic antigen-expressing tumors in human carcinoembryonic antigen transgenic mice mediated by direct lentivector injection Mol. Cancer Ther., March 1, 2009; 8(3): 692 - 702. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. P. Nishimoto, A. K. Laust, and E. L. Nelson A Human Dendritic Cell Subset Receptive to the Venezuelan Equine Encephalitis Virus-Derived Replicon Particle Constitutively Expresses IL-32 J. Immunol., September 15, 2008; 181(6): 4010 - 4018. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Escors, L. Lopes, R. Lin, J. Hiscott, S. Akira, R. J. Davis, and M. K. Collins Targeting dendritic cell signaling to regulate the response to immunization Blood, March 15, 2008; 111(6): 3050 - 3061. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Adotevi, B. Vingert, L. Freyburger, P. Shrikant, Y.-C. Lone, F. Quintin-Colonna, N. Haicheur, M. Amessou, A. Herbelin, P. Langlade-Demoyen, et al. B Subunit of Shiga Toxin-Based Vaccines Synergize with {alpha}-Galactosylceramide to Break Tolerance against Self Antigen and Elicit Antiviral Immunity J. Immunol., September 1, 2007; 179(5): 3371 - 3379. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Pichlmair, S. S. Diebold, S. Gschmeissner, Y. Takeuchi, Y. Ikeda, M. K. Collins, and C. Reis e Sousa Tubulovesicular Structures within Vesicular Stomatitis Virus G Protein-Pseudotyped Lentiviral Vector Preparations Carry DNA and Stimulate Antiviral Responses via Toll-Like Receptor 9 J. Virol., January 15, 2007; 81(2): 539 - 547. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Buffa, D. R. M. Negri, P. Leone, R. Bona, M. Borghi, I. Bacigalupo, D. Carlei, C. Sgadari, B. Ensoli, and A. Cara A single administration of lentiviral vectors expressing either full-length human immunodeficiency virus 1 (HIV-1)HXB2 Rev/Env or codon-optimized HIV-1JR-FL gp120 generates durable immune responses in mice J. Gen. Virol., June 1, 2006; 87(6): 1625 - 1634. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, RNA packaging signal; RRE, Rev response element sequences; CMVp, CMV promoter used to drive transgene expression; SV40p-Blasticidin, SV40 virus promoter used to drive selection marker blasticidin gene expression; 
U3/HIV 3'LTR, promoter deleted in U3 region so that the lvv become self-inactivated (SIN); OVA138, truncated cytoplasmic OVA gene. Restriction sites are also indicated.
