Effective immunization against neuroblastoma using double-transduced tumor cells secreting GM-CSF and interferon-gamma.

Murine neuroblastoma, neuro-2a, was transduced with the retroviral vector MFG-granulocyte-macrophage colony-stimulating factor (GM-CSF), to examine immune stimulation conferred by localized GM-CSF production. Expression of murine GM-CSF by neuro-2a (N-2a/GM) significantly reduced its tumorigenicity. Moreover, immunization of mice with irradiated N-2a/GM cells resulted in a significant protective effect against live tumor challenge 14 days later. Approximately 41% of mice immunized with irradiated N-2a/GM versus 0% of those vaccinated with irradiated parental tumor survived. Surviving mice were rechallenged after 50 days with wild-type neuro-2a or with the Sa1 syngeneic sarcoma to discern whether the generated immunity was durable and tumor specific. All mice survived wild-type neuro-2a challenge, whereas none survived inoculation with Sa1. Because both CD4+ and CD8+ T cells were necessary during priming to this MHC class Ilo, II-tumor, these data indicate that major histocompatibility complex (MHC) class I+, II+ antigen-presenting cells (APCs) were required for the T-cell antitumor response. Co-expression of GM-CSF and IFN-gamma, both of which have immunostimulatory activities on antigen-presenting cells, abrogated the tumorigenic potential of this tumor and increased immunogenicity over N-2a/IFN but not N-2a/GM. Vaccination of mice with preexisting retroperitoneal tumors with irradiated N-2a/GM and irradiated N-2a/IFN/GM improved survival. There was a trend for nonirradiated transduced cells to be more immunogenic than their irradiated counterparts. Immunohistochemistry of tissues from the vaccination site revealed a pronounced macrophage infiltration associated with nonirradiated N-2a/GM and N-2a/IFN/GM. These data suggest that vaccination involving nonirradiated neuroblastoma cells transduced with genes that stimulate APCs may be a useful approach in stimulating antitumor T-cell responses.

Irradiation of singly and doubly transduced murine neuroblastoma cells expressing B7-1 and producing interferon-gamma reduces their capacity to induce systemic immunity.

We have previously reported that immunization with low major histocompatibility complex (MHC) class I expressing murine neuroblastoma (neuro-2a) transduced with B7-1 fails to induce significant protection to wild-type tumor challenge. In this study we investigated whether B7-1 expressing neuro-2a cells can stimulate an effective T-cell response if they were cotransduced with the interferon-gamma (IFN-gamma) gene to upregulate MHC class I. Transfer of both the IFN-gamma and B7-1 genes into neuro-2a (N-2a/B7-1/IFN) almost completely abrogated the tumorigenic potential of this tumor and improved survival when compared with mice receiving the single transductants, N-2a/IFN and N-2a/B7-1. Rejection of N-2a/B7-1/IFN was mediated primarily by CD8+ T cells. When irradiated tumor cells were tested, IFN-gamma gene transfer into neuro-2a significantly increased immunogenicity, but transfer of the B7-1 gene did not. However, nonirradiated N-2a/B7-1, N-2a/IFN, and N-2a/B7-1/IFN cells were significantly more effective in eliciting systemic immunity against subsequent wild-type tumor challenge than their irradiated counterparts. N-2a/B7-1/IFN was more immunogenic than N-2a/B7-1 but not more than N-2a/IFN, indicating that B7-1 does not further increase immunogenicity of neuro-2a over that induced by IFN-gamma transduction. These findings should be considered when designing gene modified tumor vaccines for use in human trials.

B7-1 expression decreases tumorigenicity and induces partial systemic immunity to murine neuroblastoma deficient in major histocompatibility complex and costimulatory molecules.

Neuroblastoma may escape an immune attack by virtue of its low expression of surface accessory molecules essential in the antitumor response. Murine neuroblastoma, neuro-2a, was transduced with the retroviral vector LB7-1SN to examine the influence of B7-1 expression on the immune response directed against a low major histocompatibility class (MHC) I and class II negative, B7-2, and ICAM-1 negative tumor. Using a retroperitoneal model for implantation of neuroblastoma in its natural site, we demonstrated that expression of B7-1 by neuro-2a reduces its tumorigenicity. Coinjection of B7-1-positive and -negative cells improved survival compared with mice receiving B7-1-negative cells alone. This was dependent on the ratio of B7-1+ to B7-1- neuro-2a cells injected. CD8+ and not CD4+ T-cell depletion significantly increased tumor-induced mortality in syngeneic A/J mice, indicating that B7-1 decreases tumorigenicity primarily by direct constimulation of CD8+ T cells. Rejection of N-2a/B7-1 tumors or preimmunization with irradiated N-2a/B7-1 cells die not increase protection to challenge with unmodified neuro-2a cells over mice vaccinated with N-2a/neo. Furthermore, cytotoxic T lymphocyte (CTL) precursor frequencies were not significantly higher after in vivo priming and in vitro stimulation with irradiated N-2a/B7-1 compared with N-2a/neo, indicating that B7-1 costimulation by the tumor, in the absence of adequate antigen presentation by MHC molecules, may limit the generation of effective CTLs.

Short-term ex vivo activation of splenocytes with anti-CD3 plus IL-2 and infusion post-BMT into mice results in in vivo expansion of effector cells with potent anti-lymphoma activity.

We investigated the proliferation and therapeutic utility of anti-CD3 activated splenocytes infused into mice following BMT. Using congenic mouse strains we demonstrated that splenocytes activated briefly ex vivo with anti-CD3 plus IL-2 (T-activated killer cells or T-AK) and infused intravenously following BMT had a greater expansion in blood, spleen and BM compared with splenocytes stimulated with IL-2 alone. T-AK cells recovered from blood, spleen and BM consisted predominantly of CD8+ T cells. A single infusion of T-AK cells given on day +1 post-syngeneic BMT and sustained in vivo with liposomal encapsulated IL-2, significantly increased survival of mice with BDL-2 lymphoma when compared with mice receiving saline and those treated with IL-2 liposomes alone. The anti-tumor effect of T-AK cells was significantly enhanced when IL-2 was given by continuous infusion compared with bolus injections. Depletion studies confirmed that the CD8+ T-AK cells were mainly responsible for the anti-tumor effect against BDL-2 lymphoma. Our findings demonstrate that brief ex vivo activation of splenocytes with anti-CD3 plus IL-2 results in in vivo proliferation of effector cells with potent anti-tumor activity following BMT.

Retroperitoneal inoculation of murine neuroblastoma results in a reliable model for evaluation of the antitumor immune response.

To more closely mimic the natural site of human neuroblastoma and the original spontaneous arising paraspinal murine tumor, the authors developed a new model system in which murine neuroblastoma cells (neuro-2a) are implanted directly into the retroperitoneal space. This method of administration resulted in an aggressive and reproducible neuroblastoma model, with death occurring at a median of 20.3 days after tumor implantation using 1 x 10(6) neuro-2a cells, compared with the intraperitoneal (median, 31 days) and subcutaneous routes (median, 35.1 days) (P < .001). Adoptive transfer of single cell suspensions from livers, spleens, and bone marrows of mice with retroperitoneal tumors into healthy hosts resulted in tumor growth, confirming the presence of metastatic foci in these organs. The retroperitoneal murine neuroblastoma model was used to assess the importance of natural killer (NK) and T cells in regulating the growth of neuro-2a in vivo. T cells played an equally protective role as NK cells; depletion of either T or NK populations significantly decreased survival as compared with undepleted mice. Elimination of both NK and T cells further accelerated mortality of neuro-2a-bearing mice as compared to those depleted of either T or NK populations. The retroperitoneal murine model is a highly relevant in vivo system for preclinical studies of new therapeutic approaches for neuroblastoma.

Transfection of the mouse ICAM-1 gene into murine neuroblastoma enhances susceptibility to lysis, reduces in vivo tumorigenicity and decreases ICAM-2-dependent killing.

We determined the expression of intercellular adhesion molecules (ICAM) on neuro-2a cells in order to evaluate whether they were involved in cytolysis of murine neuroblastoma. Fluorescence-activated cell sorting analysis revealed that the control neomycin-resistance-gene-transduced line (neuro-2a/LN) had poor expression of ICAM-1 (mean channel fluorescence, MCF = 3.7). An ICAM-1-positive transfectant of neuro-2a (neuro-2a/ICAM-1+) (MCF = 64.3) was generated to evaluate directly the role of this adhesion molecule in cytolysis. Neuro-2a/ICAM-1+ was more sensitive to LAK killing (69.7% at an effector-to-target ratio of 100:1) compared to neuro-2a/LN (48.6%) (P < 0.001). Blocking of neuro-2a/LN and neuro-2a/ICAM-1+ lysis with anti-ICAM-1 monoclonal antibodies (mAbs) did not account for all the LFA-1-dependent killing. These data indicate that even in neuro-2a/ICAM-1+ cells, other LFA-1 ligands participated in the effector-target interaction. Therefore, we examined these cell lines for ICAM-2 expression. Both neuro-2a/LN and neuro-2a/ICAM-1+ lines expressed ICAM-2 (MCF = 16.4 and 16.5). ICAM-2 accounted for the majority of the LFA-1-dependent killing in the ICAM-1-negative target, neuro-2a/LN, while ICAM-1 played a primary role in the cytolysis of the ICAM-1+ transfectant. Inhibition of lysis in the presence of anti-ICAM-1 and ICAM-2 mAbs was comparable to that seen with the addition of anti-LFA-1 mAb, indicating that other LFA-1 ligands were not involved in this system. ICAM-1 expression was associated with decreased in vivo tumorigenicity, mice inoculated with neuro-2a/ICAM-1+ cells had a significantly longer survival compared to those receiving neuro-2a/LN cells (median survival time 35.5 versus 24.5 days) (P < 0.001). It is important to note that ICAM-1 transfection of murine neuroblastoma did not alter its metastatic potential. We conclude that transfection of mouse neuroblastoma with ICAM-1 increases its sensitivity to in vitro lysis and reduces its in vivo tumorigenicity. In ICAM-1-negative murine neuroblastoma cells, ICAM-2 plays a primary role in cell-mediated lysis.

Interleukin-2 gene transfer into murine neuroblastoma decreases tumorigenicity and enhances systemic immunity causing regression of preestablished retroperitoneal tumors.

Murine neuroblastoma, neuro-2a, was transduced with the retroviral vector LIL-2SN in order to examine the influence of localized interleukin (IL)-2 production on the immune response against a low major histocompatibility complex (MHC) class I, class II-negative, and intercellular adhesion molecule (ICAM)-1-negative tumor. Two neomycin-resistant (neo R) clones, N-2a/IL-2/L (2.5 +/- 0.4 U/ml/10(6) cells/24 h) and N-2a/IL-2/H (44.6 +/- 8.8 U/ml), were studied as representative low and high IL-2 producers, respectively. Using a recently developed retroperitoneal (r.p.) model for implantation of neuroblastoma in its natural site, we demonstrated that production of IL-2 by neuro-2a reduces its tumorigenicity in a dose-dependent fashion. T-cell, but not natural killer (NK) cell, depletion significantly increased tumor induced mortality in syngeneic A/J mice. Mice genetically devoid of T-cells (C.B-17 scid/scid) also experienced a significant increase in mortality rates. This indicates that the antitumor effect of locally secreted IL-2 is mediated primarily through activation of T-cells. Immunization of mice with irradiated N-2a/IL-2/H cells resulted in protection when challenged at a later date with unmodified neuro-2a cells. Depletion of CD8+, but not CD4+, T-cells prior to vaccination abrogated the protective effect, indicating that the priming phase of the immune response is CD8+ T-cell dependent. Mice with established r.p. tumors were vaccinated with N-2a/IL-2/H, which significantly prolonged their survival compared to unimmunized controls and to mice immunized with non-IL-2-producing neuro-2a cells. Because of the similarities of this model with the human tumor, our studies indicate that IL-2-transduced neuroblastoma cells may be effective in generating systemic immunity leading to eradication of minimal residual disease.

Importance in timing of cyclophosphamide on the enhancement of interleukin-2-induced cytolysis.

We investigated the in vivo effects of cyclophosphamide (CY) on interleukin-2(IL-2)-induced cytolytic function and spleen cell immunophenotype. Pretreatment of A/J mice with CY (25 mg/kg or 75 mg/kg) i.p. on days -10 and -15 followed by IL-2 (50,000 U i.p. on days 0 to +3) resulted in increased lysis of YAC-1 target cells compared to the group receiving IL-2 without previous CY therapy. In contrast, when CY was given on day -5, the cytotoxicity against YAC-1 was not enhanced. Phenotypic analysis of splenocytes obtained from mice treated with CY on day -10 or -15 revealed a relative decrease in L3T4- and Lyt2-positive T cells. In vivo depletion of natural killer (NK) cells by anti-asialoGM1, prior to IL-2 therapy, abrogated the enhancing effect of CY on cytolysis while in vivo elimination of T cells by anti-L3T4 and anti-Lyt2 monoclonal antibodies did not, indicating that in the absence of T cell antigenic challenge, the increased cytolytic function after CY administration is probably mediated through NK cells. These findings provide evidence that CY may be used more effectively in IL-2-based immunotherapy protocols, if consideration is given to timing of CY and IL-2 administration.