Enhanced transduction of mouse bone marrow-derived dendritic cells by repetitive infection with self-complementary adeno-associated virus 6 combined with immunostimulatory ligands

The potential of adeno-associated virus (AAV)-based vectors in human gene therapy is being explored for several diseases. Although sustained transgene expression and low vector-associated cellular immunity are attractive features of recombinant (r) AAV, the wider application of rAAV vectors encapsidated in serotype 2 capsid is hampered by poor transduction efficiency in many target tissues. These include ex vivo-generated dendritic cells (DC), which have demonstrated promising immunotherapeutic activity. We report here that efficient transduction of mouse bone marrow-derived DC can be achieved with self-complementary (sc) rAAV encapsidated in serotype 6 capsid. Sequential exposure of DC precursor cultures to IL-4 and GM-CSF with sc rAAV6 encoding the human tumor antigen, carcinoembryonic antigen (CEA), for 7 days followed by activation with CpG oligodeoxynucleotides (ODN) and anti-mouse CD40 antibody resulted in highly efficient transduction of DC. DC surface markers as determined by flow cytometry analysis of sc rAAV6-transduced DC were comparable to nontransduced DC. Efficiency of vector transduction and transgene expression were confirmed by immunostaining and real-time PCR. Microarray analysis of RNA from CpG ODN and CD40 antibody stimulated sc AAV6-transduced DC revealed upregulation of transcription factors and cytokines involved in immune activation and downregulation of inhibitory factors, suggesting a possible role of transcriptional activation in the observed effect. The adoptive transfer into syngeneic mice of the ex vivo-transduced and activated DC resulted in the development of CEA-specific antibody and T-helper 1-associated immune responses. Immunized mice also developed antibody to AAV6 capsid protein, which did not crossreact with AAV2 capsid protein. These studies demonstrate the potential utility of sc rAAV serotype 6-based vectors in transduction of DC for genetic vaccination approaches.

Coxsackievirus-adenovirus receptor genetically fused to anti-human CD40 scFv enhances adenoviral transduction of dendritic cells

A promising approach to immunotherapy involves the loading of dendritic cells (DCs) with genetic material to facilitate sustained expression of a relevant antigen in this population of potent antigen presenting cells (APC). Viral vectors such as adenovirus (Ad) have been used for this purpose. Existing methods for DC infection are limited by lack of specificity and a requirement for DC exposure to high viral doses. Targeting of Ad to DCs with bispecific antibodies has significantly augmented levels of transgene expression. Genetic fusion of the extracellular portion of coxsackievirus-adenovirus receptor (CAR) to cell-specific ligands has also proved successful in targeting Ad to cells of interest. We report here the production and primary characterization of a new fusion protein comprising the ecto-domain of CAR connected to a single chain antibody (scFv) G28-5 against human CD40 present on the surface of DCs. We demonstrate that the fusion protein (CAR/G28) specifically interacts with both recombinant Ad fiber knob and the ecto-domain of human CD40 in a binding assay (ELISA). Finally, we show that the CAR/G28 fusion protein promotes highly efficient transduction of DCs of both rhesus monkey and human origin.

Intratumoral injection of dendritic cells derived in vitro in patients with metastatic cancer

BACKGROUND

Dendritic cells (DCs) are potent initiators of immune responses, and the infiltration of DCs into tumors may confer an improved prognosis. Whether the injection of DCs directly into tumors can mediate biologic activity was examined.

METHODS

Patients with metastatic dermal or subcutaneous tumors received granulocyte-macrophage-colony stimulating factor to increase the numbers of peripheral blood monocyte precursors. DCs were then generated from monocytes obtained by phlebotomy with granulocyte-macrophage-colony stimulating factor and interleukin-4 in autologous plasma. Tumors were injected at multiple sites with 30 million autologous DCs per tumor.

RESULTS

Seven patients with melanoma and three patients with breast carcinoma were treated. Injections were well tolerated. Regression of the injected tumors, beginning as early as 4 days after injection, was observed in four patients with melanoma and in two patients with breast carcinoma. Biopsies of regressing lesions showed lymphocyte infiltration associated with DCs and necrosis. Neutrophils and macrophages were not evident. Lymphocytes expanded from the regressing tumors proliferated in response to heat shock proteins, HSP70 and gp96, derived from autologous tumor. The DCs injected produced interferon-alpha and expressed Fas ligand mRNA but did not exhibit cytolytic activity in vitro. Expression of the costimulatory molecule, B7-2 (CD86), decreased on DCs after intratumoral injection.

CONCLUSIONS

This pilot study demonstrates that DCs derived in vitro can exist viably after intratumoral injection and can mediate biologic activity in situ. Tumor-derived heat shock proteins may be involved in the antitumor activity observed.

HIV type 1-reactive chemokine-producing CD8+ and CD4+ cells expanded from infected lymph nodes

The chemokines RANTES, MIP-1alpha, and MIP-1beta have been identified as HIV-1-suppressive factors produced by CD8+ T cells. We examined the possibility that HIV-1-specific, chemokine-releasing T cells could be expanded from the lymph nodes of patients with advanced infection. Lymphocytes, separated from lymph nodes of patients with peripheral blood CD4 counts less than 500/microl obtained at diagnostic biopsies, were activated with anti-CD3 monoclonal antibody, and cultured in vitro for up to 12 days with IL-2. The phenotype, proliferative response, chemokine production, and anti-HIV-1 activity of the expanded cells was examined. Cells expanded 2.4- to 49-fold from patients with as few as 15 CD4+ cells/microl in their peripheral blood. Expanded cells were a mixture of CD8+CD45RO+ and CD4+CD45RO+ T cells. The CD8+ cells were also CD30+CDw60+CD11b-. When challenged with autologous B cell targets expressing HIV-1 Env protein, unseparated expanded cells, and purified CD8+ and CD4+ T cell subsets, proliferated and secreted MIP-1alpha and RANTES. Expanded cells were negative for HIV-1 by PCR and by culture. Culture supernatants inhibited the replication of HIV-1 in CD4+ cells in vitro. These studies indicate that HIV-1 can stimulate chemokine release by CD8+ and CD4+ cells expanded from infected lymph nodes, even from individuals with advanced infection. The numbers of chemokine-releasing T cells produced in these short-term cultures may be sufficient to be applied therapeutically as an autologous cellular therapy for HIV-1.

Human monocytes inhibit lymphokine-activated killer cell expansion in vitro

Depleting monocytes from human peripheral blood mononuclear cells (PBMC) enhances the in vitro activation of lymphokine-activated killer (LAK) cells. To determine if monocytes also altered LAK-cell expansion, we evaluated two methods of depleting monocytes from PBMC: nylon wool adherence (NWA) and phenylalanine methyl ester (PME) treatment. Both methods of depleting monocytes enhanced interleukin-2 (IL-2) driven, LAK-cell expansion; LAK expansion, however, was significantly greater after depletion with NWA than after PME. LAK cytotoxicity after NWA and PME depletion was equivalent. The degree of monocyte depletion, determined by evaluating morphology and the number of Leu-M3 (CD14) positive cells, and the proliferation of Leu 19 (CD56), OKT-3 (CD3), Leu2 (CD8), and Leu 3a (CD4) positive cells was also equivalent. Exposure of IL-2 activated cells to PME did not alter their cytotoxic activity. However, sequential treatment of PBMC with NWA, then PME, or with PME and then NWA, resulted in reduced expansion. This reduction in expansion was similar to PBMC treated with PME alone. Exposure of PME-depleted cells to nylon wool or to supernatants obtained from cells adherent to nylon wool further decreased LAK expansion relative to cells treated with NWA alone. We conclude that even at relatively low cell density, human monocytes markedly inhibit LAK-cell expansion in IL-2 driven PBMC cultures. Further, depletion of monocytes by NWA adherence is more effective than by treatment with PME, possibly due to subtle cellular damage induced by this latter treatment. These findings have implication for the in vitro and in vivo generation of LAK-cells by IL-2.

Modulation of macrophage I-A expression: lack of effect of prostaglandins and glucocorticoids on macrophages that continuously express I-A

The ability of prostaglandins, glucocorticoids and lipopolysaccharide to modulate I-A expression on macrophages that continuously or transiently express I-A was compared. We found that neither prostaglandins nor glucocorticoids affected I-A expression by macrophages that continuously expressed I-A. In contrast, prostaglandin E2 (PGE2) inhibited the lymphokine-induced maintenance of I-A on macrophages that transiently expressed I-A while glucocorticoids inhibited both the induction and expression of I-A. Lipopolysaccharide caused a loss in the expression of Class II molecules on macrophages that continuously expressed I-A by a prostaglandin-independent mechanism. The lack of effect of PGE2 and glucocorticoids may be related to our observation that continuous expression of I-A does not require its continued synthesis.

Induction of lymphokine-activated killer cells in strain 2 guinea pigs with human interleukin-2

Culture of spleen cells from strain 2 guinea pigs with Jurkat interleukin-2 (IL-2) resulted in the induction of lymphokine-activated killer (LAK) cells. Maximum LAK activity was induced using 5000 pmol of IL-2. Incubation of spleen cells with IL-2 for as little as 8 h resulted in detectable LAK activity. LAK cell activity was transient and could be stimulated by adding back IL-2. LAK cell activity was enriched in a 1.085 single-step percoll gradient. Admixture of guinea pig LAK cells with the line 10 hepatoma prior to intradermal injection resulted in inhibition of tumor growth. Systemic passive transfer of LAK cells together with concurrent IL-2 resulted in a significant inhibition of metastatic tumor growth.

Modulation of cyclic AMP-dependent protein kinase isozyme expression associated with activation of a macrophage cell line

We have used the RAW 264.7 macrophage (MO) cell line to study cAMPdPK isozymes during activation by lymphokine (LK) and lipopolysaccharide (LPS). Untreated cells were found to have two isozymes of cAMPdPK in their cytosol. PKI and PKII were differentiated based on the Mr of their regulatory subunits (RI, 45,500; and RII, 52,000, respectively) as determined by photoactivated incorporation of the cAMP analog 8-N3-[32P]cAMP. Loss of the RI subunit of PKI occurred in association with activation of the cell line by suboptimal concentrations of LK and LPS (1/40 dilution, 1 ng/ml) or high concentrations of LPS alone (10 ng/ml to 100 micrograms/ml). No modulation of the RII subunit of PKII was observed under these conditions. The loss of RI was dependent on the addition of a triggering signal to the MO. Treatment of RAW 264.7 cells with LK alone at dilutions from 1/10 to 1/1280 was not sufficient to cause a disappearance of the RI subunit from the cytosol or to induce antitumor activity. The addition of a suboptimal concentration of LPS after LK or a high dose of LPS alone was required for acquisition of cytolytic activity and loss of RI. The kinetics for the disappearance of RI from treated cells were found to be identical after activation with either LK and LPS or high concentrations of LPS alone. RI could no longer be detected in the cytosol 8 hr after the addition of activating agents. The antitumor activity of the RAW 264.7 cell line was transiently expressed after activation. Cells no longer exhibited tumoricidal activity 48 hr after the removal of activating agents. It was observed that the loss of cytolytic function was accompanied by the reexpression of RI in the cytosol. This study provides evidence that modulation of cAMPdPK isozymes occurs during activation, suggesting a potential mechanism for controlling the effects of cAMP on the MO.

Modulation of cyclic AMP-dependent protein kinase isozyme expression associated with activation of a macrophage cell line

We have used the RAW 264.7 macrophage (MO) cell line to study cAMPdPK isozymes during activation by lymphokine (LK) and lipopolysaccharide (LPS). Untreated cells were found to have two isozymes of cAMPdPK in their cytosol. PKI and PKII were differentiated based on the Mr of their regulatory subunits (RI, 45,500; and RII, 52,000, respectively) as determined by photoactivated incorporation of the cAMP analog 8-N3-[32P]cAMP. Loss of the RI subunit of PKI occurred in association with activation of the cell line by suboptimal concentrations of LK and LPS (1/40 dilution, 1 ng/ml) or high concentrations of LPS alone (10 ng/ml to 100 micrograms/ml). No modulation of the RII subunit of PKII was observed under these conditions. The loss of RI was dependent on the addition of a triggering signal to the MO. Treatment of RAW 264.7 cells with LK alone at dilutions from 1/10 to 1/1280 was not sufficient to cause a disappearance of the RI subunit from the cytosol or to induce antitumor activity. The addition of a suboptimal concentration of LPS after LK or a high dose of LPS alone was required for acquisition of cytolytic activity and loss of RI. The kinetics for the disappearance of RI from treated cells were found to be identical after activation with either LK and LPS or high concentrations of LPS alone. RI could no longer be detected in the cytosol 8 hr after the addition of activating agents. The antitumor activity of the RAW 264.7 cell line was transiently expressed after activation. Cells no longer exhibited tumoricidal activity 48 hr after the removal of activating agents. It was observed that the loss of cytolytic function was accompanied by the reexpression of RI in the cytosol. This study provides evidence that modulation of cAMPdPK isozymes occurs during activation, suggesting a potential mechanism for controlling the effects of cAMP on the MO.