Highlights of the 2nd Joint Conference of SPIRAT and NCDDG-HI

CONFERENCE COVERAGE

Note: The 2nd Joint Conference of the Strategic Program for Innovative Research on AIDS Treatment (SPIRAT) and the National Cooperative Drug Discovery Groups for the Treatment of HIV Infection (NCDDG-HIV) was held in Vienna, VA, June 22-26, 1997. Its title, "New Opportunities for HIV Therapy- From Discovery to Clinical Proof-of-Concept," is an indication of the goals of these two federal programs, which provide money to groups for the discovery and preclinical development of novel and under-explored therapies for HIV-1 infection (NCDDG-HIV) and the expedited transfer of advanced, innovative research findings to the clinical setting (SPIRAT). Thus, the information presented here was at the cutting edge of innovative but potentially clinically relevant HIV research. The following is a selected summary of that information.—Stan Deresinski, MD, FACP

HIV-1 Accessory Proteins as Potential Therapeutic Targets

Two sessions dealt with HIV accessory proteins as potential therapeutic targets. Vpr was discussed by Michael Emerman from Seattle, Michael Bukrinsky of Manhasset, and Irvin Chen of Los Angeles. The Vpr gene product is important for the infection of terminally differentiated macrophages by HIV-1 but also prevents infected cells from entering mitosis by inducing cell cycle arrest in the G2 phase. There is evidence consistent with the hypothesis that it accomplishes this G2 arrest by interfering with the DNA repair process and, as a result, with the transmission of signals necessary for cell transit from G2 to the M phase of the cell cycle. Vpr facilitates HIV DNA integration into host DNA by improving the efficiency of importation of the viral pre-integration complex to and across the nuclear pore complex. These multiple modes and sites of action of Vpr make it a potential point of therapeutic vulnerability for HIV.

Nef-deficient HIV appears to demonstrate reduced virulence, as indicated by several reported cases of natural infection as well as live vaccine experiments. Didier Trono of La Jolla, Jacek Skowronski of Cold Spring Harbor, and Kathleen Collins of Cambridge, MA, discussed the ability of the nef protein to downmodulate the expression of both CD4 and MHC I. The latter ability is consistent with the finding that HIV-1 infection of primary T cells decreases expression of class I MHC (A, B, and C alleles) 20- to 30-fold. This downregulation was dependent upon expression of the nef gene and, importantly, resulted in protection from killing by cytotoxic T lymphocytes (CTL). Thus, nef- deficient HIV may result in an infection that is more readily controlled by the host CTL response. This provides additional impetus for the targeting of nef in the preventive and therapeutic settings.

Cyclophilin A, to which the immunosuppressive agent cyclosporin binds, is a normal cell constituent that acts as a chaperone important in the folding of normal cell proteins. As discussed by Jeremy Luban of New York, however, this protein has been hijacked by HIV-1 to act as a chaperone for the gag protein and, as a result, is necessary for replication of this virus. Interestingly, if the gag-cyclophilin A interaction is disrupted by either mutations in gag or by competitive inhibition by cyclosporin A, virion assembly is not disrupted, but after entry of the virion core into a new target cell, reverse transcription is blocked. Thus, as noted by others, the gag-cyclophilin interaction may prove to be an excellent target for therapeutic intervention by cyclosporin-like drugs.

Daniel Littman of New York reviewed information about chemokine receptors and HIV pathogenesis and provided new information from studies in animal models. Of note is that there appears to be differential expression of these receptors on naïve and memory T cells. John Rossi of Duarte, CA, discussed experiments with ribozymes (RNA with the ability to catalyze the hydrolysis of complementary RNA) targeted to CCR5 alone or together with ribozymes aimed at additional targets. Another approach to interfering with CCR5 was described by Si-Yi Chen from Winston-Salem, NC. In essence, a modified chemokine ("intrakine") was targeted to the endoplasmic reticulum to block the expression of newly synthesized CCR5 at the cell surface.

Immune Cell and Gene Therapies

CD28 is a critical costimulatory molecule present on the surface of T lymphocytes. If the CD3 (T cell receptor molecule) is stimulated in the absence of costimulation, the result is either apoptosis (programmed cell death) or anergy. Bruce Levine of Bethesda discussed recent data from Carl June’s recent publication in Science demonstrating the in vitro polyclonal expansion of CD4 cells from HIV-infected subjects by stimulation with monoclonal antibodies directed at CD3 and CD28, and the resistance of these expanded CD4 cells to HIV infection.(Levine, BL. Science 1996; 278:1939-1943). This resistance to HIV infection appears to be specific to CD28 costimulation and is not observed after stimulation of other accessory receptors. In addition, this antiviral effect was specific to M-tropic (macrophage-tropic) strains of HIV and was associated with "downregulation" of the coreceptor for such strains, CCR5 (Carrol, RG. Science 1997;276:273-276). (Other data suggest that this may not represent downregulation but may instead be the result of a selective expansion of naïve [CD45RA+/CD62L+] CD4 cells, a population resistant to HIV infection and/or replication.) Of note is that CD28 costimulation also upregulates the expression of Bcl-Xl, whose gene product is anti-apoptotic. Thus, its effect should be to expand an HIV-resistant, naïve CD4 cell population, which is relatively long-lived. This group proceeded to administer expanded CD4 cells to a three HIV-infected patients in gradually increasing doses. The infusions were well tolerated and were associated with a day 7 dose-dependent mean increase of 311 + 122 CD4 cells/mcL after the highest dose of 3 ´ 104 CD4 cells. The increases were sustained for more than four months in two of the three subjects. There was, however, no apparent effect on plasma HIV RNA.

Judy Lieberman reported an update of results of her group’s study of immunotherapy with ex vivo expanded HIV-specific cytotoxic CD8+ lymphocytes. Doses of approximately 1-5 ´ 108 HIV-specific cytotoxic T lymphocytes (CTL) have been reached and were associated with a relatively transient increase in CD4 count and similarly brief decrease in plasma HIV RNA and modest decrease in cell-associated virus. Dr. Lieberman also reviewed her observation that the inability of the CTL response to eliminate HIV may be the result of down-regulation of CD3z, the signaling component of the T cell receptor, in vivo. This down-regulation, which is rapidly reversed on in vitro cultivation in the presence of IL-2, has previously been found in patients with cancer and lymphoma. The finding that the HIV nef gene down-regulates MHC1 expression (vide supra), together with this observation of decreased signaling, indicates that HIV survives by actively impairing CTL activity directed against itself.

Phil Greenberg of Seattle reported his group’s experience with the administration of autologous genetically modified CD8+ lymphocyte clones. The cells were modified to contain the Hy and HSV-TK genes; the former allows in vitro selection while the latter is a "suicide gene" which allows elimination of the clone by administration of ganciclovir or similar drugs. These cells were demonstrated to preferentially localize to lymph nodes and to co-localize to areas of infected CD4 cells within the nodes. Unfortunately, as this group has previously indicated, infused cells containing these genes were eliminated from the host because a CTL response developed to the foreign proteins produced—a very important lesson for gene therapists. Infusion of autologous clones not containing these genes are not eliminated by immunologic means, but their intravascular residence is, nonetheless, relatively brief. This is a consequence, it is believed, of a lack of CD4 help. It is now planned to co-administer, along with autologous CD8+ lymphocyte clones, autologous, genetically modified HIV-resistant CD4+ lymphocytes.

Greenberg also presented in vitro work demonstrating that a variety of genes, including RevM10, protected CD4+ lymphocytes from HIV replication and did not result in impaired function of these cells after four weeks of culture. However, in cultures maintained for 10 weeks, there was a decrease in IL-2 and IFN-g production by these cells. It was demonstrated that during such prolonged culture, these cells, all containing genes interfering with post-integration steps of HIV, became superinfected by HIV (i.e., multiple copies of HIV DNA were integrated into host cell DNA).

The HIV Rev gene product is necessary for the transport of unspliced and singly spliced mRNA from the nucleus to the cytoplasm of the infected cell. RevM10 is a mutant Rev protein that interferes with this process and thus inhibits HIV replication. Gary Nabel from Ann Arbor, MI, has also demonstrated that RevM10 inhibits CD4 cell apoptosis by preventing the formation of gp160. He reviewed his experience of autologous infusion of expanded CD4+ lymphocytes into which RevM10 has been inserted using a retrovirus vector. Preliminary results in nine patients demonstrated that cells expressing RevM10 had a four- to five-fold survival advantage relative to cells containing a non-functional control gene. Furthermore, these cells persisted and expressed RevM10 RNA for greater than 10 weeks after infusion with cell survival being apparent for up to 40 weeks. No immune response to RevM10 was detected.

Hematopoietic stem cells (HSC) are an alternative target of genes protecting progeny CD4+ cells against HIV infection or preventing its replication. Annemarie Mosely from Palo Alto, CA, described the experience with stem cell gene therapy using RevM10. A clinical trial is in progress, but insufficient follow-up is available yet to describe results. One issue being addressed in the study is the need for conditioning ("making space") prior to reinfusion of the gene-altered autologous stem cells.

Flossie Wong-Staal from La Jolla reported her group’s experience with development of an effective ribozyme gene therapy. To date, they have faced low transduction rates into HSC, and, as a consequence, the results of infusion of HSC containing a ribozyme into three subjects was disappointing. Illustrative of an impending explosion of gene therapy trials in HIV, Donald Kohn of Los Angeles described a panoply of clinical protocols involving gene altered HSC which are soon to begin.

Thymus Function and Thymic Transplantation

Immune preservation or reconstitution are key to the ultimate "cure" of HIV-infected individuals. A major stumbling block to immune reconstitution may be inadequate thymic function, which is critical to the normal maturation of T cells. Unfortunately, thymic size and function decrease with increasing age; furthermore, CD4+ thymocytes are infected by HIV, further potentially impairing thymic function in the HIV-infected individual.

Mike McCune from San Francisco examined thymic mass by CT scan in HIV-infected adults and controls. In the former group, he reported an inverse correlation between thymic mass and age and a direct correlation with CD4 count as well as number of naïve (CD45RA+CD62L+) CD4+ lymphocytes. Those less than 37 years of age with CD4 counts of 300-500 were more likely to have thymic tissue than those in lower or higher CD4 strata. Ninety percent in this group had a thymic shadow. The results, by age, were similar (or perhaps worse) in a seronegative group; among those over 40 years of age, none of 10 had a thymic shadow. Thus, thymic mass does not appear to diminish more rapidly in HIV-infected than in non-HIV-infected adults.

Nonetheless, a great deal of effort is being made in the area of thymic transplantation. Richard Hong, from Burlington, VT, reported the apparent successful transplantation of a patient with DiGeorge syndrome. He then presented a three- to eight-month follow-up of five HIV-infected patients given thymic grafts after a single dose of ATG. Their baseline CD4 counts were less than 200, and they had been on "triple therapy" for more than two months. CD4 increase was seen in three of five, and the percentage of naïve cells increased in three of four. There was no change in plasma HIV RNA. Apparently normal thymic tissue was detected on biopsy of transplant sites in two of three subjects.

M. Louise Markert of Duke University described her group’s SPIRAT thymic transplant protocol planned to enroll 16 antiretroviral naïve adults with CD4 counts of 200-500/mm3. Subjects are placed on AZT, 3TC, and ritonavir and, after six weeks, are randomized to thymic transplant or control groups. Eight patients have been enrolled, with four having received thymic grafts. Efficacy could not yet be determined because of inadequate follow-up.

Human thymic allotransplantations are generally performed using neonatal thymus obtained routinely at the time of cardiovascular surgery for repair of congenital heart defects. The supply of such tissue is, obviously, limited, making exploration of thymic xenografts an important avenue of research. Furthermore, such xenografts ought to be resistant to HIV infection, potentially making them more functional than native thymus in HIV-infected individuals. Very important work in this area is being done by David Sachs’ group at Harvard, including Megan Sykes and Michael Rosenzweig.

Megan Sykes reported that fetal miniature swine thymus tissue reconstitute mouse CD4 T cells in thymectomized, T cell depleted mice. (Neonatal thymus is less effective.) The grafts grow in size, and the recipient mouse is tolerant of pig antigens (e.g., does not reject pig skin grafts). CD4 function appears normal, and recipient mice are able to recover from experimental PCP. CD8 recovery is, however, less successful than CD4 recovery. Immune responses are host restricted. Positive thymocyte selection is porcine mediated; negative selection is mouse and pig mediated.

SCIDHuLIV+ mice given porcine fetal thymus grafts exhibit human thymopoiesis and T cell recovery. The T cells populate the mouse lymph nodes. The CD4 and CD8 lymphocytes are functional and polyclonal (by TCR Vb analysis). Pig thymus is more effective than human thymus. This, however, may be artefactual in that the pig thymus is fresh and the human thymus is one day old.

In data directly examining the potential for transplantation of porcine thymus to primates and humans, Michael Rosenzweig reported that porcine thymic stroma supports T cell differentiation of both rhesus and human CD34+ (> 90% CD34+, < 1% CD3+) cells. Resulting cells are functional and polyclonal and include NK and gamma-delta T cells, as well as CD4 and CD8 T cells.

This led to in vivo studies. A single macaque, infected with SIV239 more than two years previously, was given 20 fetal pig thymic grafts into its quadriceps without any conditioning. The animal was pretreated with trimethoprim-sulfamethoxazole and azithromycin but did not receive antiretroviral therapy until 14 days post-transplantation, when AZT and 3TC were started. There was no change in CD4 count, and viral load data are not yet available. At six weeks, the animal developed hind limb paralysis and was sacrificed. Post-mortem suggests that paralysis was due to CMV polyradiculitis + HIV spinal cord disease.

On a more successful note, David Sachs reported a return of T cells being observed in a single T cell depleted cynomolgous monkey given fetal pig thymus.

Vaccination

David Weiner of Philadelphia described the initial experience with preventive vaccination of chimpanzees with DNA vectors coding for either HIV-1 envelope protein or gag/pol proteins followed by challenge with a heterologous strain of HIV. The immunized animals have remained negative for plasma HIV RNA by PCR for one year, while a single control chimpanzee had been positive since two weeks after challenge. Studies have been initiated in humans.

Deborah Cotton reported her group’s Phase 1 study of HLA-A2.1 restricted HIV-1 (CY-2301) and HBV (CY-1899) peptides in asymptomatic subjects with HIV-1 infection. CY-2301 is a lipopeptide construct consisting of an HIV-1 CTL epitope covalently linked to a tetanus toxoid protein-derived helper T cell epitope linked to two palmitic acid moieties. It was administered to two subjects one month ago. One is now off study because of a severe local reaction.

Suppression and Ablation

The long-lived reservoir of cells latently infected with HIV present an important barrier to cure of HIV infection.

Joseph Wong and colleagues examined the effect of potent antiretroviral therapy administered for 36-52 weeks on HIV RNA and DNA in blood and lymph nodes. Ten subjects received indinavir alone, AZT and 3TC, or all three drugs. In subjects receiving triple therapy who had undetectable plasma HIV RNA (< 100 copies/mL), examination of inguinal lymph nodes demonstrated that, although the quantity of viral RNA were reduced 10,000-fold, it remained detectable. Furthermore, in those on triple therapy that had been interrupted, and in those on AZT plus 3TC alone, lymphoid viral loads were 107-109 copies per gram of tissue. In the presence of continued viral replication, drug resistance developed. These findings provide additional impetus to the need for aggressive antiretroviral therapy aimed at complete suppression of viral replication as well as the development of methods of eradicating the last reservoir of HIV-infected cells.

Various strategies are being explored to eliminate this reservoir. Clay Smith described three planned protocols in this vein. A protocol using low-dose cytoxan is in preparation in the ACTG. Two protocols are aimed initially at patients with AIDS lymphoma: CHOP (cytoxan, adriamycin, vincristine, and prednisone) therapy and a combination of total body irradiation, and myeloablative therapy followed by bone marrow transplantation. It is hoped that such therapies will not only cure the lymphoma but also eradicate HIV.

Yair Reisner described the ability to overcome major histocompatibility barriers and thus induce tolerance by transplantation of "megadoses" of HSC in sublethally irradiated mice. The transplanted cells contain tolerizing cells which, although MHC 1 and MHC 2 positive, did not possess CD80 (B7-1) or CD86 (B7-2), the ligands for the CD28 costimulatory molecule (vide supra). As a result, the investigators believe these cells presented antigen in the absence of costimulation, thus leading to anergy.

GVL in the Absence of GVHD

The graft vs. leukemia (GVL) reaction is a critical element of successful transplantation therapy of leukemias. GVL is distinct from GVHD (graft vs host disease), and a crucial element of therapy is to create GVL in the absence of GVHD. This has been successfully achieved in the treatment of acute leukemia, CML, and CLL by using nonablative preparative regimens and subsequent infusion of alloreactive cells in those with residual disease; for CML, approximately 107 cells are effective. Richard Champlin and his group are exploring the application of this approach to the treatment of HIV infection. Initial studies will involve subjects with AIDS-associated lymphoma and low CD4 counts, or subjects who have failed their initial therapy. In these patients, non-ablative preparation therapy will be followed by allotransplantation and a brief period of GVHD prophylaxis followed by infusion of alloreactive T cells. The aim is eradication not only of tumor but also of HIV-infected cells.