APOPTOSIS IN ANIMAL MODELS OF HIV-1 DISEASE

Animal models of AIDS are crucial in understanding the complexity of HIV-1 infection of suscep­tible cells throughout the body and the diverse homeostatic and immune mechanisms that respond to the loss of cells and combat the virus.

HIV-1-Infected SCID-hu Thy/Liv Mouse

The SCID-hu mouse is a small animal model created by implanting human (hu) fetal thymus and liver under the kidney capsule of a SCID mouse, resulting in a conjoint thymus/liver (Thy/Liv) organ.

The mechanisms by which HIV-1 causes depletion of thymocytes remain controversial and warrant further systematic study.

Apoptosis plays a major role in the depletion of thymocytes from HIV-1-infected Thy/Liv organs. Thymocytes with condensed nuclei, fragmented DNA, and partial chromosomal loss were observed in HIV-1-infected Thy/Liv organs but not in uninfected organs.^14,18,80 Zaitseva and col­leagues found overlapping regions of apoptotic thymocytes and immature dendritic cells in HIV- 1-infected thymus, implicating a role for DC in clearing apoptotic thymocytes.⁹⁰ Immature DP thymocytes, which express higher levels of CXCR4 and CD4 but lower levels of Bcl-2 than mature SP thymocytes, are preferentially infected and depleted by X4 HIV-1 infection of the Thy/Liv organ.⁸¹ Human thymocytes express high levels of Fas on DN and DP thymocytes but little on mature SP thymocytes.⁹¹ Cross-linking Fas with agonist MAb, however, did not cause apoptosis in human thymocytes.

Jamieson and colleagues observed few apoptotic cells at the peak of viral replication in Thy/Liv organs, suggesting a role for necrosis in HIV-1-induced thymocyte depletion.³³ Lenardo and colleagues also observed necrotic death rather than apoptotic cell death in peripheral blood mononuclear cells (PBMCs) and several T cell lines.^93,94 These results, however, should be inter­preted with caution. Apoptosis is an active process, and mitochondrial respiratory function plays a crucial role in the decision between apoptotic vs. necrotic cell death. Mitochondrial involvement was implicated in HIV-1-mediated cell death. We and others observed an increase in caspase-3 activity in HIV-1 infected thymocytes and monocytes.⁹⁵ Electron transport chain complex I was recently shown to be a substrate of caspase-3, and its destruction results in the shutdown of adenosine triphosphate (ATP) production.^96,97 This will lead to intracellular ATP exhaustion and mitochondrial rupture, causing necrotic cell death, even though cell death was originally initiated by apoptosis. We observed that treatment of thymocytes with z-VAD-FMK provides partial pro­tection against cell death induced by HIV-1. Cycloheximide treatment, however, was much more protective against cell death, suggesting that an active apoptotic process rather than a passive process is responsible for cell death in HIV-1-infected thymocytes (Shailesh K. Choudhary and David Camerini, unpublished observations).

HIV-1 Infected hu-PBL-SCID Mice

The hu-PBL-SCID mouse provides another model of HIV-1 infection, which is created by intrap­eritoneal injection of human PBMCs into SCID mice. The model was created by Mosier and colleagues in 1988, and since then it has been used extensively to study HIV-1 pathogenesis, anti­HIV therapy, and vaccine efficacy in vivo.^98-112 Human lymphocytes, including T cells, B cells, and NK cells, survive for an extended period of time, secrete human immunoglobulin, and generate secondary antibody responses in hu-PBL-SCID mice.^98,99,113 Human cells populate not only the site of peritoneal injection but also lymph nodes, spleen, and bone marrow and exhibit activated CD45RO⁺ memory populations in both CD4⁺ and CD8⁺ T cell subsets.¹¹⁴,¹¹⁵ This activated state of the T cells, therefore, mimics the lymphocyte activation seen in chronic HIV-1 infection and provides an excellent target for HIV-1 infection in hu-PBL-SCID mice.¹¹⁶ Both R5 and X4 HIV-1 can infect and deplete CD4⁺ T cells in hu-PBL-SCID mice.⁹⁹,¹⁰⁰,¹¹³,¹¹⁷ The rapid depletion of CD4⁺ T cells caused by R5 HIV-1 in hu-PBL-SCID mice was striking, because for the first time, it revealed that R5 HIV-1 could be highly cytopathic in a model system. The higher replication and CPE of R5 HIV-1 in this system was supported, in part, by increased expression of CCR5 on CD4⁺ T cells in hu-PBL-SCID mice. Hu-PBL-SCID mice engrafted with PBL from CCR5∆32∕∆32 donors were resistant to R5 HIV-1 infection, and dualtropic HIV-1 exhibited slower replication kinetics in CCR5∆32∕∆32 hu-PBL-SCID mice.¹¹⁸ Higher expression of CCR5 on CD4⁺ T cells, along with the high proportion of cells with memory phenotype, makes hu-PBL-SCID mice an excellent choice for R5 HIV-1 pathogenesis studies.^115,117 Moreover, different viral isolates exhibited differential CPEs in hu-PBL-SCID mice, reflecting the biological heterogeneity of HIV-1 observed in infected individuals.¹¹⁹ For example, Nef-deficient HIV-1 showed lower CD4⁺ depletion ability than wild­type and exhibited delayed replication kinetics.^120,121

Many modified hu-PBL-SCID mice models have recently been developed to engraft PBL better.

Although hu-PBL-SCID mice were extensively used to study HIV-1 pathogenesis, few inves­tigators focused their attention on cell death pathways after HIV-1 infection in this system. Miura and colleagues observed many apoptotic cells in the spleens of HIV-1-infected hu-PBL-NOD-SCID mice. The majority of these apoptotic cells, however, were not productively infected, indicating that bystander cell killing was the major mechanism of cell depletion in this model.¹²⁵ A combination of terminal deoxynucleotidyltransferase dUTP nick end labeling (TUNEL) assay along with immu­noassaying of TNF family members, including TNF-α, FasL, and TNF-related apoptosis-inducing ligand (TRAIL), demonstrated increased expression of TRAIL on apoptotic cells. Furthermore, admin­istration of neutralizing anti-TRAIL antibody but not anti-FasL MAb provided protection against HIV-1-induced apoptosis in hu-PBL-NOD-SCID mice, thus linking TRAIL to depletion of bystander CD4⁺ T cells.¹²⁵ Administration of lipopolysaccharides (LPSs) to HIV-1-infected hu-PBL-NOD-SCID mice induced infiltration of HIV-1-infected human cells into the brain and neuronal apoptosis.¹²⁶ Neuronal apoptosis, however, was blocked by antibodies against human TRAIL, implicating a role for TRAIL expressed on HIV-1-infected macrophages in neuronal apoptosis. Another study suggested a role for the antiapoptotic Bcl-2 protein family, because overexpression of Bcl-2 or Bcl-xL in the neuronal cell line NT-2 provided protection against HIV-1-induced apoptosis.¹²⁷ del Real and col­leagues observed increased expression of TNF-α in serum of HIV-1-infected hu-PBL-SCID mice; however, its relevance in the context of HIV-1-induced apoptosis was not investigated.¹²⁸

HIV-1-Infected Humanized Mice and Rats

The development of an HIV-1 permissive small animal model would greatly facilitate the study of HIV-1-mediated apoptosis and pathogenesis. Attempts to develop a rodent model of AIDS, however, were hampered by the inability of HIV-1 to infect primary rodent cells productively.¹²⁹,¹³⁰ The restrictions occur at various stages of the viral life cycle, including at entry, transcription, and assembly.^131-142 Goldstein and colleagues engineered mice to express human CD4 and CCR5 on their T cells (hu-CD4∕CCR5 transgenic mice) in order to remove the block to viral entry.¹⁴³ PBMCs and splenocytes from hu-CD4∕CCR5 transgenic mice expressed human CD4 and human CCR5 and were susceptible to limited infection with R5 HIV-1. HIV-1-specific Gag DNA sequences were detected in lymphocytes from the spleen and lymph nodes of hu-CD4∕CCR5 transgenic mice.¹⁴³ Entry of HIV-1 in these transgenic mouse T cells, however, did not result in production of significant numbers of viral particles. The absence of HIV-1 Tat∕Rev RNA signals in infected cells indicated a block in, or before, viral transcription in cells from hu-CD4∕CCR5 transgenic mice.¹⁴³ Similarly, Koito and colleagues developed transgenic mice expressing human CD4 and CXCR4 on their CD4⁺ T cells.¹⁴⁴ Thymocytes from these mice were susceptible to in vitro infection with HIV-1 but to a lower level than human PBMCs. Moreover, lack of in vivo data makes the hu-CD4∕CXCR4 transgenic mouse a less convincing candidate for an animal model of HIV-1 infection.

The hu-CD4∕CCR5 transgenic rat, on the other hand, may be a more promising humanized rodent model for HIV-1 infection. Primary normal rat T cells, macrophages, and microglia expressed considerable levels of early HIV-1 gene products when infected with VSV-G pseudotyped replication- competent HIV-1. Moreover, HIV-1 infection of normal rat macrophages and microglia, but not lymphocytes, produced substantial levels of HIV-1 p24 CA and infectious virions.¹⁴⁵ These encour­aging results led the Goldsmith group to develop a transgenic rat that expressed human CD4 and human CCR5 on several diverse rat cell populations.¹⁴⁶ Co-expression of hu-CD4 and hu-CCR5 was detected on rat (r) CD4⁺ T cells, macrophages, and microglia but not on r-CD8⁺ T cells or B cells. Moreover, ex vivo infection of primary macrophages or microglia from hu-CD4∕CCR5 transgenic rats resulted in productive infection and generation of replication-competent virus. In contrast, transgenic expression of human CD4 and CCR5 made rat T cells permissible for viral entry but did not result in productive HIV-1 infection. Hu-CD4∕CCR5 transgenic rats, however, can be infected with HIV-1 in vivo. Lymphoid organs from HIV-1-infected hu-CD4∕CCR5 trans­genic rats contained 2-LTR circles, integrated provirus, and early viral gene products, demonstrating susceptibility to HIV-1 in vivo. Furthermore, low levels of plasma viremia were detected in the early phase of infection, suggesting that the hu-CD4∕CCR5 transgenic rat may be a promising candidate for development into an effective small animal model of HIV-1 infection.¹⁴⁶

HIV-1 Transgenic Mice

Several investigators have developed transgenic (TG) mice containing the entire HIV-1 genome. Expression of the complete HIV-1 genome in transgenic mice led to the development of an AIDS­like disease and caused early death.^147-152 R5 HIV-1, clone JR-CSF transgenic mice had high levels of plasma viremia, which could be further increased by treatment with the superantigen Staphylo­coccus enterotoxin B (SEB) or by infection with Mycobacterium tuberculosis.¹⁵² Both T cells and monocytes from JR-CSF TG mice produced low levels of HIV-1, which could infect human PBMCs.¹⁵² Injection of JR-CSF TG mouse leukocytes into SCID-hu Thy/liv mice resulted in pro­ductive infection, indicating that these cells provided an in vivo source of replication-competent virus.¹⁵⁰ Moreover, JR-CSF TG mouse leukocytes were a persistent in vivo reservoir of HIV-1, despite treatment with HAART.¹⁵⁰ HIV-1 was also detected in microglia of these transgenic mice, which produced more MCP-1 in response to in vivo stimulation with LPS than controls.¹⁵¹ X4 HIV-1, clone NL4-3 (CD4C/HIV) transgenic mice expressed all three size classes of HIV-1 mRNAs (8.8-kb unspliced, 4.3-kb singly spliced, and 2.0-kb multiply spliced) at high levels in the thymus and moderate levels in the spleen and lymph nodes.¹⁴⁸ T cells, dendritic cells (DC), and macrophages all expressed the HIV-1 genome in these transgenic mice. The expression of full-length NL4-3 caused depletion of CD4⁺ T cells from all lymphoid organs (thymus, spleen, and lymph nodes) and caused their atrophy.¹⁴⁸ Lower expression of CD40 ligands along with reduced IL-4 production by NL4-3 transgenic CD4⁺ T cells caused impaired germinal center (GC) and follicular dendritic cell (FDC) function¹⁵³ similar to the degeneration of GC and the FDC network seen in HIV-1-infected people.⁵⁰,¹⁵⁴ Jolicoeur and colleagues observed a progressive decrease in the number of DCs in lymph nodes with selective accumulation of CDIIb^l" MHC class II^Lo HIV⁺ immature DC and a concomitant decrease in antigen presentation capacity.¹⁵³ These alterations resulted in serious immune dysfunction and development of an AIDS-like disease in NL4-3 TG mice.

After it became clear that transgenic mice expressing HIV-1 provirus in their CD4⁺ T cells, DCs, and macrophages developed AIDS-like pathology, attempts were made to delineate the HIV-1 genes responsible for lymphocyte depletion. Transgenic mice containing the HIV-1 clone NL4-3 genome lacking the Gag-Pol genes (HIV^AGag/APQl transgenic mice) had impaired growth, cachexia, and rapid progression to death, which suggested the involvement of other HIV-1 genes in the development of AIDS-like disease.^155-157 Jay and colleagues created TG mice bearing HIV-1 NL4-3 genomes with deleted Gag, Pol, and most of the Env coding sequence but retaining all the open-reading frames for accessory genes except Vpu. Profound depletion of T cells was observed in all lymphoid compart­ments, suggesting that neither virion production nor the envelope was required for T cell depletion.¹⁵⁸ Detection of apoptotic cells by TUNEL assay suggested that lymphocyte depletion in the thymus, spleen, and lymph node was the result of accelerated T cell death by apoptosis. Mature T cells in the peripheral lymphoid organs and in the thymic medulla, however, were less sensitive to apoptosis than immature T cells in the thymic cortex.¹⁵⁸ Increased expression of TNF-α was observed in sera of HIV^AGag/AM transgenic mice. Administration of anti-TNF-α MAb or human chorionic gonadotropins (hCGs) resulted in reduced TNF-α levels in sera by approximately 75% and prevented mortality, therefore implicating TNF-α in the pathology of [ IIV^λg≠'¹'"¹ transgenic mice.¹⁵⁵

Jolicoeur and colleagues and others observed that the HIV-1 Nef gene is a major disease determinant in transgenic mice.¹⁴⁸,^159-164 A fatal AIDS-like disease occurred in transgenic mice that harbored functional Nef (CD4C/HIV^Nef) but not in mice that had all HIV-1 genes except Nef.¹⁵⁹ Transgenic expression of the SIVmac239 nef gene under the control of the human CD4 promoter (CD4C/SHIV-Nef^SIV transgenic mice) also resulted in AIDS-like disease with low levels of circu­lating CD8⁺ and CD4⁺ T cells, low T cell numbers in the thymus and peripheral lymphoid organs, thymic atrophy, splenomegaly, weight loss, wasting, and premature death.¹⁶⁵ Peripheral TG CD4⁺ T cells from CD4C/HIV^Nef transgenic mice exhibited an activated/memory phenotype with higher expression of CD25 and CD69 on CD44^high, CD45RB^low, CD62L^low, and CD4⁺ T cells, even in the absence of TCR-mediated stimulation.¹⁶² Furthermore, thymocytes from these transgenic mice showed hypersensitivity to stimulation with CD3 MAb; constitutive increases in tyrosine phospho­rylation of several substrates, including LAT and p44/p42 MAPK (Erk-1/Erk-2), were also observed in these transgenic mice.¹⁵⁹ The chronic activation of transgenic CD4⁺ T cells and their hypersen­sitivity to TCR/CD3-mediated signals caused apoptosis after limited cell division.¹⁶² This contrib­uted to exhaustion of the T cell pool, thymic atrophy, and the low numbers of CD4⁺ T cells observed in CD4C/HIV^Nef transgenic mice. The importance of Nef in this model is evident, because mutation in the SH3 ligand-binding domain of Nef resulted in abrogation of hypersensitivity to CD3 stim­ulation, and mice bearing this mutant Nef failed to develop disease, despite partial downmodulation of CD4 cell surface expression.¹⁶¹

HIV^at transgenic mice that express Tat under the control of the astrocyte-specific glial fibrillary acidic protein promoter in a doxycycline (Dox)-dependent manner were developed by He and colleagues to study neuropathological disorders associated with HIV-1 infection.¹⁶⁶ The expression of Tat occurred exclusively in astrocytes and was Dox dependent. Expression of Tat in the brain caused damage to the cerebellum and cortex, astrocytosis, degeneration of neuronal dendrites, neuronal apoptosis, and increased infiltration of activated monocytes and T lymphocytes similar to what is seen in the brains of AIDS patients.¹⁶⁶ HIV^at transgenic mice, therefore, support the hypothesized role of Tat protein in HIV-1 neuropathogenesis and may be helpful in delineating the molecular mechanisms of Tat-induced neurotoxicity and apoptosis.¹⁶⁶

The HIV^Vpr transgenic mouse was developed by Yasuda and colleagues to investigate CPE mediated by Vpr in vivo.¹⁶¹ The X4 HIV-1 clone NL4-3 Vpr was expressed under the control of the mouse CD4 enhancer/promoter in transgenic mice. HIV^Vpr transgenic mice had reduced T cell numbers in both the thymus and peripheral blood due to enhanced apoptosis. The highest-level expression of Vpr was observed in the thymus followed by lymph node (50-fold lower) and spleen (100-fold lower), which was correlated with tissue phenotype; the thymus exhibited atrophy, whereas the spleen and lymph node appeared to be normal in HIV^Vpr transgenic mice. Crossing the HIV^pr transgenic mice with Gld-mice, which have deficient FasL genes, still caused thymic atrophy, suggesting that FasL is not required for apoptosis induced by Vpr. Increased expression of Bcl-x, Bax, and caspase-1 genes, however, was observed in thymuses of these transgenic mice, suggesting the involvement of mitochondrial pathways in Vpr-mediated apoptosis.

HIV-1-mediated apoptosis was implicated in neuronal damage, resulting in cognitive and motor dysfunction, known as HIV-1-associated dementia.¹⁶⁸ HIV-1 gp120 was demonstrated to promote apoptosis directly on isolated neurons.^169,170 HIV-1 gp120 (HIV^gp120) transgenic mice exhibited a 25% reduction in CNS dendritic cells compared with wild-type littermates.¹⁷¹ Administration of caspase inhibitors prevented dendritic cell damage associated with gp120 expression in HIV^gp120 transgenic mice. Moreover, crossing HIV-1 gp120 transgenic mice with Casp-1∕C285G transgenic mice, which have reduced production of mature IL-1β in brain, caused dendritic cell numbers to return to the number found in wild-type littermates. This suggests that dendritic cell injury in HIV-1 gp120 transgenic mice is due to inhibition of caspase-1 or blockade of mature IL-1β production.¹⁷¹ Transgenic mice, therefore, provide a novel in vivo model with which to systematically investigate the signals and pathways associated with HIV-1 apoptosis.

HIV-1-Infected Chimpanzees

Besides humans, chimpanzees are the only species that can be infected with HIV-1.^172-176 Chimpan­zees were experimentally infected with various HIV-1 isolates, but infection did not normally result in AIDS or other HIV-1-associated pathogenesis. One exception, however, was animal C499. This animal was part of a superinfection study and was infected with three different HIV isolates: HIV- ¹SF2, HIV-1_la∖⅛ and HIV_NDK.¹⁷⁷ The animal developed an AIDS-like syndrome with severe CD4⁺ T cell loss with a concomitant increase in viral load, thrombocytopenia, and chronic diarrhea associated with cryptosporidium infection.¹⁷⁸ Blood transfusion from C499 to an HIV-1-negative chimpanzee (C455) resulted in rapid CD4⁺ T cell loss in the recipient animal, suggesting that in vivo evolution of a chimpanzee-pathogenic HIV-1 strain occurred.¹⁷⁸ HIV-1 isolation from chimps C499 and C455 and subsequent molecular cloning suggested that virulent clones were likely recombinants derived from HIV-1_sf2 and HIV-1_lav.¹⁷⁹ Infection of other chimps with C499 HIV-1 isolates resulted in moderate to high viral load in plasma and peripheral CD4⁺ T cell depletion.^180-182 None of the viral isolates derived from chimp C499, however, caused AIDS-like disease in experimentally infected chimpanzees, even after years of infection.^183,184 Attention, therefore, focuses on the potential mech­anisms for the lack of disease progression in HIV-1-infected chimps. Gougeon and colleagues observed that HIV-1 infection of chimpanzees did not result in chronic activation of T cells, in contrast to HIV-1-infected humans.¹⁸⁵ CD4⁺ and CD8⁺ T cells from infected chimpanzees were less susceptible to apoptosis mediated by HIV-1 or TCR ligation.^185-189 Furthermore, Fas ligation by agonistic antibodies or recombinant human Fas ligand failed to induce apoptosis on CD4⁺ or CD8⁺ T cells from HIV-1-infected chimpanzees.¹⁸⁵ These observations could be partly explained by the fact that chimpanzee T cells are resistant to FasL upregulation and oxidative stress.¹⁸⁸ T cells from HIV-1-infected chimpanzees exhibited normal responses to antigen and TCR-mediated stimula­tion.¹⁸⁷ Chimpanzees that exhibited signs of progression to AIDS, however, had increased levels of activated T cells compared to those chimps that did not progress, suggesting that chronic T cell activation is a primary determinant of disease progression in HIV-1-infected chimpanzees.^181,182

SHIV-Infected Macaques

Simian-human immunodeficiency viruses (SHIVs) are chimeric viruses in which env and the associated auxiliary genes tat, rev, and vpu of SIVmac are replaced with the corresponding regions of HIV-1 to develop a better primate model of HIV-1 infection.^190,191 Some SHIVs, including SHIV89.6P, are pathogenic in infected macaques, causing progressive loss of CD4⁺ T cells and development of an AIDS-like disease.^192-194 Infection of macaques with pathogenic SHIVs, there­fore, may provide insights into the mechanisms of CD4⁺ T cell depletion in HIV-1-infected people.

Infection of macaques with pathogenic SHIV strains caused apoptosis in both CD4⁺ and CD8⁺ T cell populations. In contrast, nonpathogenic SHIV strains were inefficient inducers of apoptosis in rhesus macaque PBMCs.^195-198 Similarly, pathogenic SHIVs caused increased FasL expression in both CD4⁺ and CD8⁺ T cells, whereas nonpathogenic SHIVs were conspicuous by their inability to increase FasL expression on PBMCs of infected macaques.^197,199 These results suggest that induction of apoptosis is closely associated with the pathogenic process induced by SHIV in macaques. Apoptotic cells were detected in the lymph nodes and the thymuses of infected macaques.^197,199 An in situ TUNEL assay showed increased numbers of apoptotic cells in the paracortical regions of lymph nodes and medulla of thymus, which colocalized to a large extent, with sites of active viral replication and CD4⁺ T cell depletion.²⁰⁰ Pathogenic SHIV infection caused increased expression of activation markers on macaque CD4⁺ T cells, including HLA-DR, CD25, and CD69.²⁰¹ Expression of CD95L was higher on HLA-DR⁺ T cells than HLA-DR T cells.¹⁹⁹ Activation of PBMCs with PHA caused substantial increases in apoptosis of T cells, predominantly among CD4⁺ T cells.²⁰¹ Furthermore, T cells from blood and lymphoid tissue did not proliferate in response to PHA, suggesting the generation of dysfunctional T cells.²⁰¹ These data are consistent with previous obser­vation that apoptosis correlated with immune system activation and higher expression of CD95, CD95L, TNF, and IFN-γ.^{56,199,202,203} The chronic activation of the immune system in SHIV-infected macaques, therefore, causes an increased rate of T cell destruction and impedes immune system regeneration, which may be responsible for the development of simian AIDS.