Gene therapy offers a radical different approach to the treatment of arthritis. Here we have demonstrated that two marker genes (lacZ and neo) and cDNA coding for a potentially therapeutic protein (human interleukin 1-receptor-antagonist protein; IRAP or IL-lra) can be delivered, by ex vivo techniques, to the synovial lining of joints; intraarticular expression of IRAP inhibited intraarticular responses to interleukin 1. To achieve this, lapine synoviocytes were first transduced in culture by retroviral infection. The genetically modified synovial cells were then transplanted by intraarticular I jection into the knee joints of rabbits, where they efficiently colonized the synovium. Assay of joint lavages confirmed the in vivo expression of biologically active human IRAP. With allografted cells, IRAP expression was lost by 12 days after transfer. In contrast, autografted synoviocytes continued to express IRAP for -5 weeks. Knee joints expressing human IRAP were protected from the leukocytosis that otherwise follows the intraarticular injection of recombinant human interleukin 1p3. Thus, we report the intraarticular expression and activity of a potentially therapeutic protein by genetransfer technology; these experiments demonstrate the feasibility of treating arthritis and other joint disorders with gene therapy.Arthritis is a chronic, debilitating condition affecting over 30 million Americans (1). Presently incurable, it remains the agent of considerable suffering and economic loss. Therapeutic intervention in arthritis is hindered by a number of factors, including difficulties in targeting drugs to joints. Proteins are particularly vulnerable to this restriction, which is of special concern, as many new agents with considerable antiarthritic potential are proteins. As an alternative to traditional methods of drug delivery, we have suggested the transfer oftherapeutic genes to the synovial lining ofjoints (2,3). Expression of these genes would overcome proteindelivery problems and lead to the intraarticular accumulation of the gene products at the site of disease, with reduced exposure of nontarget organs.Using the rabbit knee joint as a model system, we are therefore developing in vivo and ex vivo methods for delivering genes to joints. This model takes advantage of the similarity in size between the knee joint of the rabbit and the human proximal interphalangeal joint, a common site of rheumatoid arthritis. Moreover, there exists a large body of literature on the biology of the rabbit's knee, including well-established methods for synovial cell culture (e.g., refs. 4-6). Here we report the transfer to synovium of two marker genes and one potentially therapeutic gene by an ex vivo approach. Intraarticular expression of the interleukin 1-re-The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. (rhIL-1,8). These results demonstrate the feasibility of ...

Objective. To assess the abilities of various vectors to transfer genes to the synovial lining of joints.Methods. Vectors derived from retrovirus, adenovirus, and herpes simplex virus as well as cationic liposomes and naked plasmid DNA were evaluated. Each construct contained the lac Z marker gene; and one retroviral construct, and one plasmid also contained a gene encoding human interleukin-1 receptor antagonist. Gene expression was under the control of the human cytomegalovirus promoter in all vectors except the retrovirus, where the endogenous 5' long terminal repeat was retained as the promoter. Cultures of rabbit synovial fibroblasts were exposed to these vectors and stained with X-gal to identify lac Z+ cells. Vectors were then injected directly into rabbits' knee joints, and gene transfer and expression were assessed by X-gal staining and polymerase chain reaction (PCR).Results. Adenovirus was a highly effective vector both in vitro and in vivo, with lac Z gene expression persisting for at least 28 days. However, an inflammatory response was noted in vivo. Cells infected in vitro and in vivo with herpes simplex virus also expressed the loc 2 gene at high levels, but expression was limited by cytotoxicity. Retroviruses, in contrast, were effective --~ _-

Dendritic cells (DC) have been successfully used in clinical pilot studies to induce tumor-specific immunity as well as clinical response in selected patients. However, DC-based immunotherapy remains a challenge and several parameters need to be examined in order to optimize the induction of anti-tumor immune responses. This study focuses on DC vaccination for leukemia and evaluates the in vitro efficacy of three different strategies for generating antigen-loaded DC-based vaccines for the induction of major histocompatibility complex (MHC) class I-restricted anti-leukemia cytotoxic T lymphocyte (CTL) responses. These included direct fusion of DC with leukemia cells to generate DC-leukemia cell hybrids, and DC pulsed with either apoptotic leukemia cell fragments or whole tumor cell lysates. Using either the U937 cell line or primary human acute myeloid leukemia blasts (AML), DC-leukemia cell hybrids were found to be the most potent in vitro inducers of CTL activity. DC pulsed with apoptotic tumor cell fragments were less efficient, but induced a more potent CTL response compared to tumor lysate-pulsed DC. The CTL responses were both MHC class I-restricted and antigen-specific, as shown by the inability of the CTL to lyse other control targets. The data presented here suggest that the method of antigen loading onto DC may be critical in the design of tumor vaccines.

Three strategies have been designed to concentrate infectious retroviral vectors from the supernatants of human- (HT1080) and murine- (NIH 3T3) based packaging cells. Streptavidin-conjugated paramagnetic particles in conjunction with (i) antibodies directed against murine fibronectin, (ii) biotinylated lectins, or (iii) biotin-modified packaging cell-surface proteins allow affinity-mediated magnetic concentration of retroviral vectors. Retroviral titers (assayed by colony formation of human myeloid K562 cells) are increased by 1-4 x 10(3)-fold after volume reductions of only 125-fold. Using these procedures, preparations of 5 x 10(8) cfu/ml are routinely made from relatively low-titer (2-5 x 10(5) cfu/ml) starting material. High-titer (paramagnetic) retroviral vector preparations can be used for magnetic field-dependent retroviral infection in vitro. Magnetic field-dependent localization such as this may enable the in vivo administration of formulations that concentrate retroviral infection to the required target tissues and organs.

Dendritic cells (DCs) primed with tumor antigens (Ags) can stimulate tumor rejection. This study was aimed at evaluating the polarization of T-cell responses using various DC Ag-priming strategies for vaccination purposes. DCs cocultured with irradiated "apoptotic" tumor cells, DC-tumor fusions, and DCs pulsed with freeze-thaw tumor lysate Ags served as Ag-primed DCs, with EG7 tumor cells (class II negative) expressing OVA as the model Ag. DCs loaded with class I- and class II-restricted OVA synthetic peptides served as controls. Primed DCs were assessed by the in vitro activation of B3Z OVA-specific CD8 T cells and the proliferation of OVA-specific CD8 and CD4 T cells from OT-I and OT-II TCR transgenic mice, respectively. In vivo responses were measured by tumor regression following treatment with Ag-primed DCs and by CTL assays. Quantification of IL-2, IL-4, IL-5, IFN-gamma, and TNF-alpha by cytometric bead array (CBA) assay determined the polarization of TH1/TH2 responses, whereas H-2 Kb/SIINFEKL tetramers monitored the expansion of OVA-specific T cells. DC-EG7 hybrids stimulated both efficient class I and class II OVA responses, showing that DC-tumor hybrids are also capable of class II cross-presentation. The hybrids also induced the most potent CTLs, offered the highest protection against established EG7 tumors and also induced the highest stimulation of IFN-gamma and TNF-alpha production. DCs cocultured with irradiated EG7 were also effective at inducing OVA-specific responses, however with slightly reduced potency to those evoked by the hybrids. DCs loaded with lysates Ags were much less efficient at stimulating any of the OVA-specific T-cell responses, showed very little antitumor protection, and stimulated a weak TH1 response, overbalanced by an IL-5 TH2 response. The strategy of Ag-loading clearly influences the ability of DCs to polarize T cells for a TH1/TH2 response and thus determines the outcome of the elicited immune response, during various vaccination protocols.