Development of a mouse model for HIV/AIDS.

Abstract

A small animal model would be very valuable for HIV/AIDS vaccine testing, investigating HIV pathophysiology, and exploring anti-HIV therapeutics. Unfortunately, HIV does not replicate in mouse cells. Provision of mouse cells with human CD4, CCR5 and cyclin TI (cycT1) has uncovered a block to HIV assembly or release. Since mouse-human cell fusions allow viral replication, mouse cells lack at least one critical factor that permits completion of the viral life cycle. To identify this factor(s) we are employing 2 similar genetic approaches. Each cell line of a panel of monochromosomal mouse-human somatic cell hybrids was individually transduced with an HIV vector encoding both cycT1 and blasticidin resistance (HIVCIB). Each was then transfected with vesicular stomatitis virus (VSV) G protein and measurable virus was recovered from only the hybrid-containing chromosome 2. This was verified with an M-tropic envelope and was shown to be specific to HIV. In addition, the amount of p24 release from that hybrid was substantially greater than that from the parent. A second cell line expressing chromosome 2 had a similar phenotype. CycT1 has been introduced into one chromosome 2 fine to monitor the spread of HIV. In a related but separate approach, an entire collection ~500 mouse-human microcell hybrids was transduced with HIVCIB and broken down into manageable pools. Virus was similarly recovered as above from a few of the pools. Those pools were then broken down to clones and several cell clones have been identified that allow virus release. Revertants that no longer have the human chromosome are now being tested for loss of phenotype. Clones will then be tested for ability to support both HIV replication and Gag processing. Human chromosomal content of the clones of greatest interest will be determined by STS content analysis. Results from the 2 approaches are expected to be in agreement and may provide direction for an expression cloning approach.

Presentation Summary

The final presentation of the afternoon, given by Richard E. Sutton, MD, PhD, was about the development of a mouse model for HIV/AIDS. Sutton first outlined some of the reasons why a small-animal model of HIV disease would be beneficial. He admitted that while such a model might not be as useful in therapeutics development, there could be considerable benefit for vaccine testing and pathogenesis research. Also, the animals used in the nonhuman primate model are expensive and scarce, thus preventing some research from moving forward more quickly. A mouse model of HIV/AIDS would allow the application of transgenesis techniques and the initial testing of many candidate vaccines that have not yet been tested because of a shortage of rhesus monkeys. In addition, mouse studies have already resulted in some discoveries about HIV, including the identification of co-receptors, cyclin T1 (a necessary co-factor for Tat and TAR during transcriptional elongation), and host factors required for viral assembly and release. Mouse cells have been modified to express CD4, a co-receptor, and cyclin T1, and the processes of HIV entry, integration, and transcription can be reproduced in mice. However, virus release from cells has not been achieved. HIV-infected mouse cells exhibit gag precursors (p55) and little capsid. Fusion of mouse with human cells causes a marked increase in viral production, suggesting mouse cells lack certain factors that are needed to release virus. For instance, viral protease may require a co-factor found in human cells or the Gag protein may be incorrectly ubiquitinated in mouse cells. As Sutton noted, the possible explanations abound for why this occurs, and there are 2 approaches to solving the problem: biochemical and genetic. Sutton's lab has chosen the latter.

The group used mouse-human somatic cell hybrids to screen a panel of mouse cell lines each containing a single human chromosome. The hybrids containing human chromosome 2 produced markedly higher levels of HIV than the other monochromosomal hybrids. Other tests show that this characteristic carried on chromosome 2 appears to be specifically associated with HIV release. Also, the cells containing chromosome 2 had levels of p24 and p17 comparable to human cells, and supernatant from the cell cultures expressing chromosome 2 contained viruses that were able to infect cells at a similar frequency as supernatant from infected human cells. Mouse cells without human chromosome 2 did not show these effects.

Another approach used by Sutton's group involves a collection of microcell hybrids where each clone was derived from a mouse melanoma cell line and contains a small amount of a single human chromosome. An HIV-based vector encoding cycT1 was introduced into the entire collection. The clones were separated into smaller pools and media from each clone was tested for viral infectivity of human cells. Certain cell clones produced very good HIV release. Ongoing preliminary analysis has shown that some of these clones contain part of chromosome 2. The lab will also look at whether reverting the phenotype in these clones will cause the cells to lose the ability to release infectious virus.

Sutton and his colleagues have also begun to employ expression-cloning strategies using an HIV cDNA expression vector. This strategy allows the introduction of a cDNA library into non-dividing cells and subsequent completion of complex functional selections. The group plans to use this technology for several investigations including the identification of unknown viral receptors as well as human host factors that are critical for HIV replication. Initial work has involved the introduction of the cDNA library into mouse cells containing cycT1. The cDNA and vector were rescued by transfecting the mouse cells with packaging vector and vesicular stomatitis virus G (VSV G). The released virus was then amplified in human cells--only as an intermediate to increase the levels of virus, which were very low in the mouse cells. The virus was recovered from the human cells and reintroduced into mouse cells. This process was repeated several times to enrich for vectors encoding cDNA that allow completion of the viral lifecycle (as seen with the cell hybrids containing chromosome 2; see Figure). The amount of virus recovered increased after each round of this process, indicating the enrichment of some factor possibly responsible for improving virion release. However, cDNAs recovered at this point were considered artefactual and not relevant to the HIV life cycle. Therefore, the cDNA vector used in this set of experiments was not considered optimal and an improved vector is already being tested.

[FIGURE OMITTED]

Sutton admits that using the human cells as an intermediate step for amplification may not have been ideal, and future studies will look at ways to improve the selection strategy (without human cells). The group is considering a standard, albeit laborious, cDNA screening approach. If such a virus-release factor was isolated, one goal would be to create a quadruply-transgenic mouse to study HIV pathogenesis and vaccine development. Another objective would be to learn how the putative factor might increase particle infectivity. One application of this work may be the eventual identification of a novel therapeutic target.

Further Reading

Mariani R, Rutter G, Harris ME, Hope TJ, Krausslich HG, Landau NR. A block to human immunodeficiency virus type 1 assembly in murine cells.J Virol. 2000;74:3859-3870.

Bieniasz PD, Cullen BR. Multiple blocks to human immunodeficiency virus type 1 replication in rodent cells. J Virol. 2000;74:9868-9877.

Reed M, Mariani R, Sheppard L, Pekrun K, Landau NR, Soong NW. Chimeric human immunodeficiency virus type 1 containing murine leukemia virus matrix assembles in murine cells.J Virol. 2002;76:436-443.

Richard E. Sutton, MD, PhD, Ayse K. Coskun, MD, and Van Nguyen

Baylor College of Medicine