Molecular modification of idiotypes from B-cell lymphomas for expression in mature dendritic cells as a strategy to induce tumor-reactive CD4⁺ and CD8⁺ T-cell responses

B-cell lymphomas are characterized by the clonal expansion of tumor cells expressing an idiotypic immunoglobulin (Id). The Id is determined by rearrangements of the variable heavy (V_H) and light (V_L) chains of the immunoglobulin V regions that are unique for each clonal B-cell population and represent tumor-specific antigens. The Id contains epitopes recognized by antibodies^1-3  and epitopes presented to T cells.^2,4-10  These unique features have led to the development of several strategies for active and passive Id-specific immunotherapy.^11,12  Initial clinical trials demonstrated that passive transfer of monoclonal Id-specific antibodies can be effective in the treatment of non-Hodgkin B-cell lymphoma (NHL),¹  although tumor escape often limits the efficacy of this approach.^13-16  Active immunization with tumor-derived Id protein was first shown to be effective in mouse models of lymphoma,^2,3,17  and clinical trials of vaccination with patient-specific Id protein have been shown to elicit immunologic anti-Id and clinical responses.⁸  Initial trials, with the Id protein coupled to keyhole limpet hemocyanin (KLH), induced mainly humoral responses to the antigen, with robust cytotoxic cellular responses only in a minority of patients. T cells have been shown in animal models and clinical trials to mediate potent antitumor effects. The importance of CD8⁺ T cells in antitumor responses has been often emphasized,^18,19  and infusion of tumor-reactive CD8⁺ T cells expanded to large numbers in vitro has been effective in eliminating tumors in animal models and patients with cancer.^20-29  CD4⁺ T cells can not only provide critical support for the generation and maintenance of CD8⁺ responses but also mediate antitumor activity.^30,31  Thus, efforts have been made to boost CD4⁺ and CD8⁺ responses to the Id. Strategies have included fusion of the antigen with xenogeneic constant regions, or fusion of a single-chain variable fragment (scFv) with molecules to enhance immunogenicity, such as fragment C of tetanus toxoid,³²  cytokines such as granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin 4 (IL-4), IL-2, IL-1β.^33-36  Clinical trials produced encouraging immunologic cellular responses that correlated with improvement in relapse-free and overall survival. Following administration of Id-KLH conjugates with GM-CSF, Id-specific cellular responses were induced in 15 of 23 patients, and immune responses were associated with molecular remission in most patients.³⁷  Administration of dendritic cells (DCs) pulsed with Id induced cellular Id-specific immune responses in 65% of patients.^9,27  Despite the observed clinical benefit, the efficacy of these vaccines may have been limited in part by the inefficient induction of Id-specific CD8⁺ T-cell responses, presumably as a result of providing the antigen exogenously as a protein. Accordingly, ongoing clinical trials evaluating DNA vaccination with scFV coupled with fragment C are producing improved CD8⁺ responses.¹²  To elucidate the importance and relative contributions of Id-specific CD4⁺ and CD8⁺ T cells in the control of B-cell malignancies, we have performed studies in the 38C13 murine model for human NHL.¹⁷  To achieve antigen processing in the context of major histocompatibility complex (MHC) class I, the Id was inserted into a recombinant vaccinia virus (rVV) for infection of DCs.^38,39  As we have previously demonstrated, vaccinia infection blocks DC maturation,⁴⁰  making it essential that DCs be matured prior to infection for induction of CD8⁺ responses. However, following maturation, DCs exhibit severely reduced ability to capture and process antigens for presentation in the context of MHC class II for the generation of CD4⁺ T cells. Therefore, we and others have used strategies that directly target intracellular antigens to the MHC class II compartments while still permitting class I presentation.^40-44  In particular, the cytoplasmic tail of the lysosomal-associated membrane protein 1 (LAMP1) has been shown to be sufficient for lysosomal targeting.⁴⁵  By using rVV containing the Id from the 38C13 murine B-cell lymphoma fused to LAMP1-targeting signals, we demonstrated that DCs can elicit Id-specific CD4⁺ and CD8⁺ T cells capable of protecting animals from tumor challenge and curing tumor-bearing mice.

Six-week-old female C3H/HeN mice (Charles River, Calco, Italy) were housed in a specific pathogen-free animal facility and treated in accordance with the European Union guidelines and with the approval of the Institutional Ethical Committee.

38C13 is a carcinogen-induced murine B-cell lymphoma tumor that expresses IgM/k on its surface and is MHC I⁺, MHC II^-.^46,47  All experiments were performed with cells from a cell bank of frozen 38C13 cells. 38C13 cells were maintained in RPMI 1640, 10% heat-inactivated fetal calf serum (FCS), 2 mM l-glutamine, 100 U/mL penicillin, 100 μg/mL streptomycin, and 50 μM 2-Mercaptoethanol (complete medium) at 37°C, 5% CO₂ in a humidified incubator. CH27 is a carcinogen-induced B-cell lymphoma tumor that also expresses IgM/k on its surface, but this idiotype is different from the one expressed by 38C13. CH27 cells were maintained under the same culture conditions as 38C13.

DCs were prepared from bone marrow of C3H mice. Briefly, single-cell suspensions of bone marrow were seeded into 6-well plates at 2 × 10⁶/mL in Iscoves Modified Dulbecco Medium (Cambrex, Verviers, Belgium) supplemented with penicillin-streptomycin and 10% FCS, GM-CSF (25 ng/mL), and IL-4 (5 ng/mL; BD, San Diego, CA) for 5 days. Eight hours before harvesting the nonadherent and loosely adherent cells, lipopolysaccharide (LPS; 1 μg/mL) was added to the culture medium as a maturation signal for DCs.

The purity of DCs was evaluated by cell surface staining with phycoerythrin (PE)–conjugated antibodies (Abs) to CD11c and PE-conjugated isotype controls (BD Biosciences, Erembodegem, Belgium). Activation of DCs was evaluated by staining the cells with PE-conjugated Abs to CD80, CD86, and I-A^K (BD Biosciences) and isotype controls. The samples were analyzed on a FACSCalibur (BD Biosciences) using CellQuest software. Data shown are representative of 3 or more independent experiments.

The Sig-HA1-V_L-LAMP1 and Sig-V_H-HA1-LAMP1 genes were constructed by 20 cycles of strand overlap extension polymerase chain reaction (PCR) with the high-fidelity PWO DNA polymerase (Roche, Mannheim, Germany), using as template the Sig-E7-LAMP1 gene (provided by D. Pardoll, Johns Hopkins University, Baltimore, MD)⁴⁵  and the V_L and V_H sequences from the 38C13 cDNA. The genes contain the 5′ sorting signal of the mouse LAMP1 (Sig) (to mediate translocation of the nascent antigen to the lumen of the endoplasmic reticulum) encoded in frame with HA1-V_L or V_H-HA1 (where HA1 is a tag-peptide derived from hemagglutinin of influenza virus) and the 3′ transmembrane and cytoplasmic tail of the mouse LAMP1 (for lysosomal targeting of the antigens). Fusion genes were constructed as shown in Table 1. HA1 was placed at 5′ of V_L and 3′ of V_H to facilitate cloning. The rVV-Sig-HA1-V_L-LAMP1, rVV-Sig-V_H-HA1-LAMP1, rVV-HA1-V_L, and rVV-V_H-HA1 were generated by cloning the genes in the pSC11T7 plasmid downstream of the vaccinia p7.5 early/late promoter by sticky/blunt or sticky/sticky ligation in the specific enzyme restriction sites. The plasmids were transfected with Lipofectamine (Gibco, Grand Island, NY) into BSC-40 cells infected with wild-type vaccinia (rVV-wt), and rVV was amplified in thymidine kinase–negative cells and plaque purified. An rVV coding for green fluorescence protein (GFP) under control of the vaccinia p7.5 promoter was provided by L. Corey (Fred Hutchinson Cancer Research Center, Seattle, WA). As a negative control of in vivo experiments, we used an rVV containing the S65L chimeric gene, encoding for the immunodominant pp65 protein (65) of cytomegalovirus, fused to the leader sequence (S), and lysosomal-targeting signal (L) of LAMP1.⁴⁰  Viral stocks were prepared in BSC-40 cells, titrated simultaneously, and diluted in Dulbecco modified Eagle medium (DMEM) containing 2.5% FCS to a final titer of 10⁹ plaque-forming units/mL. Chimeric constructs were also cloned in pGEM4Z/GFP/A64 (kindly provided by E. Gilboa and S. Nair, Duke University Medical Center, Durham, NC), under the T7 promoter, using as restriction enzymes BamHI and EcoRI, present in the 3′ and 5′ of the fusion genes and in the polylinker of the plasmid. These plasmids were used for in vitro transcription of mRNA with the T7 mMessage mMachine kit (Ambion, Houston, TX) and polyadenylated with Poly(A) enzyme (USB, Cleveland, OH) according to manufacturers' instructions.

To verify the intracellular localization of transgene expression, DCs were infected for 8 hours with the different rVVs, exposed to UV light for 4 minutes, washed 3 times in phosphate-buffered saline (PBS) with Ca⁺⁺ and Mg⁺⁺, and fixed on poly-l-Lysine–coated cover slips. DCs were permeabilized by exposure to saponine 0.1%, bovine serum albumin (BSA) 0.3% for 30 minutes at 20°C and stained with fluorescein isothiocyanate (FITC)–-conjugated anti-HA1 (BabCO, Richmond, CA) and PE-conjugated anti–I-A^k antibodies. Colocalization was determined by confocal microscopy (Leica Alembic, Milan, Italy). To measure IL-12 production, DCs were treated with Brefeldin A (10 μg/mL; 5 hours at 37°C), labeled with biotin-conjugated anti-CD11c mAb followed by cyanine 5 (Cy5)–PE-streptavidin, fixed in 4% paraformaldehyde, permeabilized with 0.1% saponin–1% FCS-PBS, incubated with PE-conjugated mAbs specific for IL-12 for 20 minutes at room temperature, and analyzed by FACScan.

Mice were immunized with 0.3 × 10⁶ syngeneic DCs infected with recombinant vaccinia viruses coding for Id, Id-LAMP1,or S65L as negative control.⁴⁰  Cells were injected subcutaneously in 200 μL 0.9% NaCl.

Fourteen days after vaccination, animals were killed, and spleens and lymph nodes were harvested. Splenocytes were resuspended in complete medium at 2.5 × 10⁶/mL in T25 flasks with 1 × 10⁶ 30 cGy irradiated 38C13 or CH27 cells. After 4 days, medium was supplemented with 25 IU/mL IL-2. Twelve hours later, cells were isolated on a Lympholyte-M gradient (Cedarlane, Hornby, ON, Canada) and positively selected based on CD8 expression (Miltenyi Biotech, Bergisch Gladbach, Germany) according to manufactures' instructions. At day 5, CD8⁺ selected cells were tested for cytolytic activity in a standard 4-hour ⁵¹Cr release assay at increasing effector-target (E/T) ratios in triplicates. Specific lysis was expressed according to the following formula: 100 × (average experimental cpm - average spontaneous cpm)/(average maximum cpm - average spontaneous cpm).

Splenocytes were stimulated with 30 cGy irradiated 38C13 or CH27 tumor cells as described “In vitro cytotoxicity against tumor cell lines,” and at a day 4 IL-2 (25 UI/mL) was added. After a total of 8 days of culture, cells were isolated on a Lympholyte-M gradient, and CD4⁺ cells were purified by negative selection with anti-CD8–conjugated beads (Miltenyi Biotech). To test for proliferation to the idiotype, DCs generated from naive C3H mice were transfected with mRNA coding for Sig-HA1-V_L-LAMP1 and Sig-V_H-HA1-LAMP1 or S65L as a control, as previously described.⁴⁸  Briefly, 1 × 10⁶ DCs were transfected with 5 μg RNA previously mixed with transmessager (Qiagen, Milan, Italy), at an RNA/lipid ratio of 1:4. Transfection was performed at 37°C for 12 hours. Transfected DCs were then matured with LPS (1 mg/mL) for 8 hours and used as stimulators. The CD4⁺ responders were plated with stimulators at increasing responder-stimulator (R/S) ratio in triplicates. After 72 hours, 1 μCi (0.037 MBq) [³H]thymidine deoxyribose ([³H]-TdR) was added to each well, and counts per minute (cpm) were counted at 96 hours. The proliferative response was expressed as cpm sample - cpm control. Proliferation experiments were repeated after exposure of DCs to chloroquine 50 μM immediately prior to infection with rVV Id-LAMP1 and rVV-S65L.

38C13 cells were grown in tissue culture for 72 hours, and cells in logarithmic growth phase were washed 3 times in PBS and appropriately diluted for injection of mice subcutaneously with 1000 viable tumor cells in 0.1 mL PBS. For adoptive therapy experiments, cyclophosphamide (50 mg/kg) was administered intraperitoneally 7 days after the tumor inoculum to reduce the tumor burden. Mice were followed for survival and tumor growth.

Seven C3H mice per group were immunized once with DCs infected with the different rVVs. After 2 weeks, mice were killed, and spleens and draining lymph nodes were collected and pooled within each group, depleted of B cells, and positively selected for CD4⁺ and CD8⁺ T cells. Immune selected cells (6 × 10⁶) were resuspended in 200 μL 0.9% NaCl and injected intravenously in 6 groups of 5 C3H mice that received 1000 viable 38C13 tumor cells subcutaneously and cyclophosphamide intravenously 9 and 7 days earlier. As controls, 2 cohorts of mice received only tumor cells or tumor cells and cyclophosphamide. Mice were again followed for survival and tumor growth.

Monoclonal antibodies (mAbs) GK1.5 (rat IgG2b anti–murine CD4) and 53.6.72 (rat IgG2a anti–murine CD8) (Abinova, Taiwan) were used for in vivo depletion of CD4⁺ and CD8⁺ T-cell subsets, respectively. Groups of 6 mice were each injected intraperitoneally with either 300 μL anti–CD4 (GK1.5) or anti–CD8 (53.6.72) mAbs or with both anti-CD4 and anti-CD8 mAbs at day -3, -1, +2, and +5, with day 0 designated as the day of vaccination with DCs infected with the different rVVs. Six nondepleted mice were analyzed as controls. At day 14 mice were challenged with 1000 viable 38C13 tumor cells subcutaneously and followed for survival. On the days of vaccination (day 0) and tumor challenge (day 14), blood samples were collected and analyzed by flow cytometry for CD4⁺ and CD8⁺ T-cell subsets.

Data obtained from DC maturation experiments were analyzed by a 2-tailed Student t test (Excel; Microsoft, Redman, WA). Survival proportions were analyzed by a Fisher Exact test (Prism; GraphPad, San Diego, CA). The results were considered to be significant when P was less than .05.

Infection of human immature DCs with rVV blocks the ability of DCs to mature in response to maturation signals.^40,49  Since DC maturation is required for stimulation of T-cell responses, and since many viral activities that interfere with antigen presentation are species specific, we tested whether this blockade also occurred in murine DCs. Bone marrow–derived immature DCs obtained from C3H mice were infected for 8 hours with a rVV encoding GFP (rVV-GFP) at varying multiplicities of infection (MOI). An MOI of 50:1 reproducibly resulted in the highest infection efficiency (10%-40%) and was used for subsequent experiments. Eight hours after infection, DCs were exposed to LPS and analyzed after an additional 24 hours. As previously observed with human DCs, rVV infection of murine immature DCs failed to induce up-regulation of MHC class II, costimulatory molecules, or IL-12 secretion (Figure 1A). The infected DCs could not subsequently be matured by exposure to LPS as demonstrated again by failure to up-regulate I-A^k, CD86, and weak induction of IL-12 (Figure 1B). Therefore, we examined whether murine DCs already matured by exposure to LPS remained, similarly to human DCs, susceptible to vaccinia virus infection. Immature DCs were exposed to LPS for 8 hours and then infected with rVV-GFP at an MOI of 50:1 for 8 hours. Mature DCs were susceptible to rVV infection and maintained an activated phenotype after infection (Figure 1C-D). Expressions of I-A^k (Figure 1E) and CD86 (Figure 1F) were significantly higher (P < .05) in rVV–infected mature DCs (exposed to LPS for 8 hours before infection) than in rVV-infected immature DCs subsequently exposed to LPS for 8 hours. Expressions of I-A^k and of CD86 were similar in uninfected mature DCs and in mature DCs subsequently infected with rVV. Additionally, mature DCs were more resistant to the cytopathic effects of rVV than immature DCs (data not shown). Since DCs have been reported to be more effective when infused 8 hours rather than 48 hours after exposure to an activation signal,⁵⁰  in vivo experiments were performed with DCs after 8-hour exposure to LPS followed by rVV infection.

To target antigen processing into the class II presentation pathway, rVVs coding for the variable regions of the heavy and light chain of the 38C13-Id fused with LAMP1-targeting signals were constructed. Two chimeric constructs containing the V_H or V_L region of 38C13-Id fused with the leader sequence of LAMP1, with an HA1 tag to monitor protein expression and with the cytoplasmic tail of LAMP1 were constructed and named rVV-Sig-V_H-HA1-LAMP1, rVV-Sig-HA1-V_L-LAMP1. Two chimeric constructs containing the V_H or V_L regions fused with the marker HA1 (rVV-V_HH-HA1, rVV-HA1-V_L) but lacking the LAMP1 targeting sequences were also constructed and used as controls (Table 1). To accurately compare the different constructs, viral titers were measured simultaneously 3 times. The same levels of antigen expression were detected by Western blot on large cell lines (LCLs) infected with the different rVVs (data not shown). Intracellular trafficking of V_H and V_L was assessed in rVV-infected DCs by immunofluorescence. DCs infected with rVV-V_H-HA1 and rVV-HA1-V_L showed a homogeneous pattern of antigen expression dispersed throughout the cytosol (Figure 2), whereas DCs infected with rVV-V_H-HA1-LAMP1 and with rVV-HA1-V_L-LAMP1 showed a vesicular pattern of antigen expression. This vesicular pattern and overlapping staining with anti-IA^k is consistent with localization to the endosomal/lysosomal pathway.⁵¹ 

Bone marrow–derived DCs obtained from C3H mice were activated with LPS, divided into 4 samples, and infected with the 4 rVVs: rVV-Sig-V_H-HA1-LAMP1, rVV-Sig-HA1-V_L-LAMP1, rVV-V_H-HA1, and rVV-HA1-V_L. Eight hours after infection samples were UV-irradiated to halt viral replication. Mature DCs infected with rVV-Sig-V_H-HA1-LAMP1 and mature DCs infected with rVV-Sig-HA1-V_L-LAMP1 were mixed at 1:1 ratio to obtain the Id-LAMP1 vaccine and inoculated subcutaneously into 4 mice at a total dose of 0.3 × 10⁶ DCs/mouse. Mature DCs infected with rVV-V_H-HA1 and mature DCs infected with rVV-HA1-V_L, mixed at 1:1 ratio to obtain the Id vaccine, were inoculated at the same dose into 4 other mice. Finally, mature DCs infected with the irrelevant rVV-S65L were infused into 4 mice as controls. Fourteen days after vaccination, mice were killed, and spleens and lymph nodes were harvested and analyzed in vitro. No differences in the total number of lymphocytes or CD4⁺ or CD8⁺ T cells were observed in samples obtained in mice that received different vaccines. To identify and quantify Id-specific T-cell responses, splenocytes were first stimulated with irradiated 38C13 tumor cells or CH27, as negative control, for 5 days as described in “Materials and methods.” After one round of stimulation, CH27 had not induced cell growth, while 38C13 induced proliferation for both CD4⁺ and CD8⁺ effectors. Id-specific CD4⁺- and CD8⁺-mediated immune responses were analyzed separately. CD4⁺ T cells were isolated as described in “Materials and methods,” with more than 85% purity achieved. Since 38C13 is MHC II negative, we used DCs transfected with mRNA coding for Id-LAMP1 or S65L as a negative control to detect proliferation of Id-specific CD4⁺ cells in a 72-hour ³H-thymidine incorporation assay. This sequence of stimulator cells was used to permit detection of responses to Id epitopes without confounding responses to vaccinia. A strong Id-specific CD4⁺ T-cell proliferative response was elicited in cells from mice immunized in vivo with the Id-LAMP1 vaccine and stimulated in vitro with the Id-LAMP1 constructs, with very little proliferation detected in cells harvested from mice vaccinated with Id or S65L (Figure 3A). However, mice primed in vivo with S65L did demonstrate an appropriate proliferative response to pp65 (Figure 3B). To directly demonstrate that LAMP1 sequences mediate MHC class II processing of the idiotype, chloroquine was used to block lysosomal activity. Proliferation of CD4⁺ cells from immunized mice to DCs infected with rVV-Id-LAMP1 (Figure 3C) or rVV-S65L (Figure 3D) was abrogated by exposure to chloroquine, while antigen processing in the context of MHC class I was unaffected (not shown). To assess Id-specific CD8⁺ T-cell responses, 25 × 10⁶ splenocytes were stimulated with 1 × 10⁶ 38C13 cells irradiated with 30 cGy. After 4 days, CD8 effectors were positively selected and recultured for an additional 24 hours in medium supplemented with 25 IU/mL human recombinant IL-2. At day 5, CD8⁺-selected cells were evaluated for cytotoxic activity by measuring lysis of the 38C13 cell line. All animals vaccinated with Id-based vaccines developed a tumor-specific lytic response (Figure 3D), whereas animals vaccinated with the irrelevant rVV failed to lyse 38C13. Id-specific lysis was significantly higher by cells from Id-LAMP1–vaccinated mice than from mice vaccinated with Id. The syngeneic isotype-matched CH27 cell line, which expresses a different idiotype, was not lysed, documenting specificity of the Id response (Figure 3E). Thus, Id-specific CD4⁺ and CD8⁺ T-cell responses were elicited by a vaccine expressing a modified Id in mature DCs. The magnitudes of CD8⁺ and more significantly of CD4⁺ T-cell responses were significantly enhanced if the antigen was targeted to the lysosomal/endosomal compartment.

To determine whether the immune responses elicited by this vaccine were able to protect animals from tumor development, groups of 12 mice were vaccinated with mature DCs infected with rVV-Id, rVV-Id-LAMP1, and rVV-S65L, as previously described. Fourteen days after vaccination, mice were challenged with 1000 viable 38C13 cells subcutaneously and followed for survival and tumor growth. Animals with tumors larger than 10 mm were killed, in accordance with our institutional guidelines. All animals vaccinated with the control S65L vaccine developed progressive tumors, requiring killing within 25 days of tumor challenge (Figure 4). In mice vaccinated with the Id-based vaccine 35% survived tumor free for 60 days. By contrast, all mice vaccinated with the Id-LAMP1–based vaccine demonstrated delayed tumor growth and prolonged survival, with 73% being tumor free at 60 days. Protection in animals vaccinated with DCs infected with rVV-Id-LAMP1 was significantly higher than in animals vaccinated with DCs infected with rVV-Id (P < .05) and in animals vaccinated with DCs infected with rVV-S65L (P < .001).

To directly evaluate the therapeutic role of the Id-specific T cells elicited by our vaccines and to identify which T-cell subset is required in the effector phase of the immune response to an established 38C13 tumor, we performed adoptive transfer experiments with Id-specific T cells in tumor-bearing mice. Mice were first challenged with 1000 viable 38C13 tumor cells subcutaneously. After 7 days, mice received cyclophosphamide (50 mg/kg) to reduce tumor burden and delay tumor growth.¹⁷  Forty-eight hours later, T cells from mice previously vaccinated with Id-LAMP1 or S65L vaccines were isolated as described in “Materials and methods,” and 6 × 10⁶ CD4⁺ or 6 × 10⁶ CD8⁺, or 6 × 10⁶ CD4⁺ + 6 × 10⁶ CD8⁺ T cells/mouse were adoptively transferred into tumor-bearing mice (Figure 5). Eighty percent of animals that received CD4⁺ and CD8⁺ T lymphocytes and 60% of animals that received only CD8⁺ T cells obtained from mice vaccinated with Id-LAMP1–based vaccines were completely cured and remained tumor free for up to 60 days, while none of the mice adoptively transferred with only CD4⁺ T cells survived tumor free (P < .05). T cells obtained from animals vaccinated with control S65L-based vaccines demonstrated no antitumor activity. These results demonstrate that Id-specific CD8⁺ T cells, in the absence of B cells, play a crucial role in the control of this lymphoma. Although CD4⁺ T cells exhibited no significant therapeutic activity alone, cotransfer of CD4⁺ T cells with CD8⁺ T cells enhanced the activity of the CD8⁺ T cells, suggesting either a helper effect or some direct antitumor activity against this MHC class II^- tumor.

To evaluate the role of CD4⁺ and CD8⁺ T cells in the priming phase of the Id-specific immune response, mice were depleted with anti-CD4 and/or anti-CD8 antibodies at days -3, -1, +2, and +5, as described in “Material and methods.” Mice were vaccinated at day 0 with mature DCs infected with rVV-Id-LAMP1, the irrelevant rVV-S65L, or with rVV-Id. Since we previously showed that rVV-Id fails to induce Id-specific CD4⁺ cells, analysis of antitumor response in this group of mice was used to discriminate the potential role of Id-specific CD4⁺ T cells (induced only by LAMP1-based vaccines) and CD4⁺ cells specific for vaccinia-derived epitopes (induced both by Id-LAMP1 and Id vaccines) that might be providing help in the induction of an effective immune response to 38C13. Fourteen days after vaccination, mice were challenged with 1000 viable 38C13 tumor cells subcutaneously and followed for survival. Before vaccination (day 0) and before tumor challenge (day 14) blood samples were collected, and the efficacy of T-cell depletion (> 99% in both cases) was confirmed by flow cytometry. Mice depleted of both CD4⁺ and CD8⁺ T cells or depleted of only CD8⁺ cells developed progressive tumor within 12 to 13 days (Figure 6A), confirming that CD8⁺ T cells are crucial for protection from tumor growth. In the group of Id-LAMP1–vaccinated mice, 73% of nondepleted mice survived tumor free for 60 days as compared with only 33% tumor-free survivors in CD4⁺-depleted mice, suggesting that CD4⁺ T cells likely play an important role in the priming phase of the immune response. This finding was further supported by analysis of mice vaccinated with rVV-Id (Figure 6B), in which no significant difference in tumor-free survival was observed between the group of nondepleted mice and mice depleted of CD4⁺ cells (tumor-free survival of 28% and 17%, respectively, at 60 days). Moreover, depletion of CD4⁺ T cells mitigated the protective advantage of the Id-LAMP1 vaccine compared with the Id vaccine (tumor-free survival of 33% versus 20%). Thus Id-specific CD4⁺ T cells appear to play a beneficial role in the induction phase of an effective immune response to this lymphoma.

Results demonstrate that protective CD4⁺ and CD8⁺ T-cell responses to a lymphoma idiotype can be induced by immunization with DCs expressing a modified antigen. In the past decade, the treatment of NHL has dramatically changed with the use of monoclonal antibodies, particularly, anti-CD20.⁵²  Although clearly highly effective, passive mAb therapy usually induces transient responses, suggesting the need to incorporate additional therapeutic strategies.⁵³  One potentially attractive approach has been to induce T-cell responses to the lymphoma, since the cellular arm of the immune system remains largely intact during the B-cell depletion associated with anti-CD20 therapy. Since each B-cell lymphoma expresses an accessible individually unique tumor antigen encoded by the idiotype, current vaccine trials for NHL have attempted to enhance T-cell anti-Id–specific immune responses. This goal has been complicated by the difficulty in isolating candidate MHC class I–binding epitopes in the complementarity-determining regions (CDRs) or in mutated framework region (FWR) from the different idiotypes. However, published data strongly suggest that T-cell responses can be elicited to human lymphoma idiotypes.^27,54-57  A variety of approaches are currently being explored, including vaccination with recombinant Id proteins, Id-encoding adenoviruses, Id-pulsed DCs, and DNA vaccination.^7-9,37,54  Most of these strategies have been tested in the 38C13 model, a challenging and potentially predictive model for immunotherapy of NHL.^2,46,47,55  Although Id-specific CD8⁺ and/or CD4⁺ T-cell responses have been elicited in different lymphoma models.^58,59  Direct evidence of Id-specific T cells has been difficult to reveal in the 38C13 model.^55,60  In this study, we show that, with appropriate antigen presentation, Id-specific CD4⁺ and CD8⁺ T cells can be induced in the 38C13 model. Several strategies were used to enhance immunogenicity, including expression of the Id protein directly in mature dendritic cells by way of recombinant vaccinia viruses, and the use of the sorting signal of LAMP1 to directly target the introduced idiotype protein into the endosomal-lysosomal compartment to facilitate MHC class II antigen processing.

The ability of an attached LAMP1 cytoplasmic tail to promote intracellular class II processing of an endogenously expressed protein has been described by several groups examining several different antigens. Lin et al⁴⁵  initially demonstrated the role of the lysosomal-targeting sequence with the E7 antigen of the human papilloma virus in enhancing class II presentation, and Rowell et al⁶¹  reported similar results with the HIV-1 envelope protein. We also previously published studies demonstrating that it is possible to induce simultaneous stimulation of CMV-specific CD4⁺ and CD8⁺ T-cell responses by fusing LAMP1 with the immunodominant pp65 protein from CMV.⁴⁰  Alternative targeting sequences, such as the invariant chain-targeting signal, have also been shown to promote MHC class II antigen processing of endogenously expressed proteins. However, Bonehill et al⁶²  recently showed that the LAMP1 sequence is more effective than the invariant chain-targeting signal for stimulating a CD4 T-cell response to the tumor antigen melanoma-associated antigen 3 (MAGE3). All together these data document the efficiency of the LAMP1 sequence in enhancing antigen presentation in the context of MHC class II molecules. Effective induction of T-cell responses requires that the antigen be presented by mature rather than immature DCs, which is likely not entirely achieved with many in vivo antigen delivery systems unless the antigen is provided directly in mature DCs.^38,39  In accordance with observations we previously made with human DCs,⁴⁰  murine mature DCs infected in vitro with rVV were found to maintain a functional mature phenotype. Thus, rVV infection of mature DCs maximizes antigen processing and presentation by the most potent antigen-presenting cells. Mature DCs infected with rVV-Id-LAMP1 were found to efficiently present Id-derived epitopes in the context of both MHC II and MHC I, inducing both CD4⁺ and CD8⁺ Id-specific T responses after only one round of vaccination. The Id-LAMP1 fusion vaccine was both qualitatively and quantitatively better than the Id vaccine, not only eliciting stronger CD8⁺ responses but also uniquely inducing CD4⁺ responses to the Id protein. The enhanced MHC class II processing and induction of CD4 responses represented the desired goal of the introduction of targeting sequences, but the increased level of Id-specific CD8⁺ responses induced by Id-LAMP1 was less expected. Potentially, this result could reflect an indirect consequence of the increased CD4⁺ helper response and/or a direct consequence of increased protein instability and degradation for the class I pathway from improper folding due to the fusion.⁴⁰  The separate contribution of the 38C13 V_H and V_L regions in the induction of the Id-specific immune response is still a matter of investigation. Initial experiments performed with mice vaccinated with DCs expressing only Sig-HA1-V_L-LAMP1 or Sig-V_H-HA1-LAMP1 suggest that V_H epitopes are targets of CD8⁺ effectors, while V_L epitopes are required for the generation of a CD4⁺ response (not shown). Accordingly, search for candidate H2-Kk–binding peptides, performed by the BioInformatics and Molecular Analysis Section (BIMAS, Bethesda, MD) software revealed a predicted epitope for the V_H region. The isolation of relevant epitopes for cellular immune responses will be important steps for the future. Results of this study have provided the unique opportunity to evaluate and dissect the role of Id-specific CD4⁺ and CD8⁺ T cells in 38C13 tumor control. Indirect evidence of antitumor T-cell activity was provided by the correlation between the stronger cellular response and protection from tumor growth observed in animals vaccinated with the Id-LAMP1 compared with animals vaccinated with Id. To dissect the role of Id-specific CD4⁺ and CD8⁺ T cells, we performed 2 different sets of experiments. Adoptive transfer experiments demonstrated that CD8⁺ Id-specific T cells are the critical cells for the effector phase of the response, and depletion experiments demonstrated that Id-specific CD4⁺ T cells contribute during the induction phase of immune response. It is intriguing to note that CD4⁺ depletion did not prevent activation of Id-specific CD8⁺ cells in our model. This may be secondary to the use of mature DCs, licensed antigen-presenting cells able to bypass the need for CD4⁺ T cells for the generation a primary CD8⁺ response.^63-65  Since most B-NHLs are MHC II⁺, Id-specific CD4⁺ T cells might be expected to play a potentially more important direct effector role in this clinical setting than observed in the 38C13 model. Tumor protection and cure were obtained in a high proportion of mice compared with results already reported in the 38C13 model,^55,66  suggesting that a strong T-cell response may significantly improve the treatment of NHL.

We thank David Maloney, Anna Mondino, and Catia Traversari for helpful discussion.