Pathogenic organisms are frequently attenuated after long-term culture in vitro. The mechanisms of the attenuation process are not clear, but probably involve mutations of functions required for replication and pathogenicity in vivo. To identify these functions, a direct comparison must be made between attenuated genomes and those that remain pathogenic in vivo. In this study, we used the heterochimeric SCID-hu Thy/Liv mouse as an in vivo model to define human immunodeficiency virus type 1 (HIV-1) determinants which are uniquely required for replication in vivo. The Lai/IIIB isolate and its associated infectious molecular clones (e.g., HXB2) were found to infect T cell lines but failed to replicate in the SCID-hu Thy/Liv model. When a lab worker was accidentally infected by Lai/IIIB, however, HIV-1 was isolated only from infection of primary PBMC, and not from infection of T cell lines. We hypothesized that the lab worker was exposed to a heterogeneous viral stock which had been attenuated by passage in immortalized T cell lines. Either a rare family member from this stock was selected for in vivo replication or, alternatively, an attenuated genotype dominant in vitro may have reverted to become more infectious in vivo. To address this hypothesis, we have used the SCID-hu Thy/Liv model to study the replication of HXB2 and of HXB2 recombinant viruses with HIV-1 fragments isolated from the infected lab worker. HXB2 showed no or very low levels of replication in the Thy/Liv organ. Replacement of its subgenomic fragment encoding the envelope gene with a corresponding fragment from the lab worker isolate generated a recombinant virus (HXB2/LW) which replicated actively in SCID-hu mice. The NEF mutation in the HXB2 genome is still present in HXB2/LW. Thus, the LW sequences encode HIV-1 determinants which enhance HIV replication in vivo in a NEF-independent mechanism. The specific determinants have been mapped to the V1-V3 regions of the HIV-1 genome. Six unique mutations in the V3 loop region of HXB2/LW have been identified which contribute to the increased replication in vivo.

No abstract

No abstract

The development of effective therapies for the treatment of AIDS would be facilitated by a better understanding of HIV pathogenesis in vivo. While some aspects of pathogenesis may be assessed by standard tissue culture assays, in vivo animal models may provide clues to other aspects of HIV-mediated progression toward AIDS. Current animal models include primate models for the study of simian immunodeficiency virus (SIV) and HIV, SCID-hu and hu-PBL SCID mouse models for the study of HIV, and feline models for the study of feline immunodeficiency virus (FIV). In general these models are costly and labor intensive. We have developed a simple human fetal thymic organ culture (TOC) system that is permissive for HIV infection and that exhibits pathology similar to that observed in vivo. A key feature of this system is the time-dependent destruction of thymocytes typified by the preferential loss of CD4-expressing cells. HIV-mediated thymocyte destruction occurs by a process involving programmed cell death. We have infected TOC with a panel of HIV isolates and found that the resulting viral replicative and pathogenic profiles are similar to those seen in the SCID-hu Thy/Liv mouse, yet different from profiles observed in standard PHA-blast tissue culture assays. In addition, we find that TOC may be used to assess efficacy of antiviral agents such as AZT (3'-azido-3'-deoxythymidine) and ddI (2',3'-dideoxyinosine) in blocking both viral replication and virus-induced pathology. These results indicate that this model is amenable to the systematic manipulation, analysis, and characterization of a variety of HIV virus isolates and antiviral therapies.

The scaffold attachment region of the human interferon beta gene (IFN-SAR) inserted into a retroviral vector improved transgene expression in human primary CD4+ and CD8+ T cells, and in primary monocytemacrophages. In T cells, expression of the Maloney murine leukemia virus (Mo-MuLV)-based retroviral vectors was high in activated cells but low in resting cells. Addition of the IFN-SAR sequence enhanced vector expression 2- to 10-fold, and the effect was particularly pronounced in resting T cells. In CD33+CD14+CD4+ monocyte-macrophages derived from transduced hematopoietic stem/progenitor cells (HSPCs) in vitro, the IFN-SAR enhanced vector expression three- to sixfold. We have used the IFN-SAR-containing vectors to express the RevM10 gene, a trans-dominant mutant of the human immunodeficiency virus type 1 (HIV-1) rev gene. Compared with a standard retroviral vector, the IFN-SAR-containing vector was significantly (p < 0.01) more potent at inhibiting HIV-1 replication in infected CD4+ peripheral blood lymphocytes. In monocytes, however, addition of the IFN-SAR did not significantly improve antiviral efficacy. To understand better the reason for the strong effect of the SAR on antiviral efficacy in T cells we have studied the expression of HIV, Mo-MuLV, and Mo-MuLV + SAR vectors in resting and activated cells. While the expression of all three vectors was lower in resting compared with activated cells, the kinetics of the decrease in expression were fastest for the Mo-MuLV vector, followed by the HIV vector and then the Mo-MuLV + SAR vector. Thus, higher level expression of the Mo-MuLV + SAR vector relative to wild-type HIV at all stages of T cell activation is the most likely explanation for the strong antiviral efficacy. Overall, this study demonstrates the utility of the IFN-SAR sequence for achieving high-level retroviral vector expression in lymphoid and myeloid hematopoietic cells.