HIV-1 and drug abuse comorbidity: Lessons learned from the animal models of NeuroHIV

Abstract

Various research studies that have investigated the association between HIV infection and addiction underpin the role of various drugs of abuse in impairing immunological and non-immunological pathways of the host system, ultimately leading to augmentation of HIV infection and disease progression. These studies have included both in vitro and in vivo animal models wherein investigators have assessed the effects of various drugs on several disease parameters to decipher the impact of drugs on both HIV infection and progression of HIV-associated neurocognitive disorders (HAND). However, given the inherent limitations in the existing animal models of HAND, these investigations only recapitulated specific aspects of the disease but not the complex human syndrome. Despite the inability of HIV to infect rodents over the last 30 years, multiple strategies have been employed to develop several rodent models of HAND. While none of these models can accurately mimic the overall pathophysiology of HAND, they serve the purpose of modeling some unique aspects of HAND. This review provides an overview of various animal models used in the field and a careful evaluation of methodological strengths and limitations inherent in both the model systems and study designs to understand better how the various animal models complement one another.

1. Introduction

Despite the ability of combination antiretroviral therapy (cART) to dramatically suppress viremia, the brain continues to be a reservoir for low-level HIV-1 replication [1-6]. It has been estimated that ~30-60% of infected individuals on the cART regimen go on to develop a varying degree of neurological dysfunction ranging from asymptomatic to mild neurocognitive impairments, collectively termed as HIV-associated neurocognitive disorders (HAND) or NeuroHIV, that significantly affects the quality of life of infected individuals [7-12]. Adding further complexity to HAND is the co-morbidity of drug abuse [13-17]. The Centers for Disease Control and Prevention define HIV-1 infection and drug abuse as intertwined epidemics, leading to compromised cART adherence and consequent exacerbation of HAND. The interplay of HIV-1 and drug abuse raises concern regarding the combinatorial effects of both on HAND progression [17-20]. Chronic low-level inflammation, mediated likely by residual viral proteins, cART, and drugs of abuse, has been implicated as a significant factor underlying the progression of HAND. More importantly, neuroinflammation has been recognized as an essential component of HAND.

Various studies conducted on the association between HIV-1 infection and drug addiction clearly demonstrate the role of various drugs of abuse impairing both the immunological and non-immunological pathways of the host, ultimately augmenting HIV-1 infection and disease progression [21–32]. Based on the epidemiologic studies, several animal models were employed to validate findings in the animal models of HAND. However, due to the long incubation period for the development of NeuroHIV, these investigations used several transitional outcome measures, such as CD4+ T cell decline, functional immune markers, HIV-1 load, and constitutional symptoms of HIV-1 infection, as well as neurological manifestations of HIV-1 infection. While the nonhuman primate is the gold standard model of HIV-1 infection, it is cost-prohibitive. Furthermore, HIV-1 does not infect rodents and felines – together, and these limitations have hampered adequate recapitulation of NeuroHIV.

Although several excellent non-human primate models of disease have provided valuable information on the disease process [33–37], working with macaques poses many other challenges, such as the need for a large team of investigators and infrastructure and the limitation of numbers of animals available for the study design (cost-driven). As a result, there is a need for developing alternative models of HAND. Over the last 30 years, several unique strategies have been employed for developing rodent models of HAND. While none of these models accurately mimic the human syndrome, they serve to model some aspects of HAND. These rodent models include: 1) Severe combined immunodeficient model involving the direct stereotactic injection of HIV-1-infected human monocytes into the CNS of immunodeficient mice [38–43]; 2) Direct injection of HIV proteins either into the CNS or via intravenous route [44–49]; 3) Transgenic mice with constitutive or inducible expression of HIV proteins [50–55]; 4) HIV Transgenic rat model expressing all of the HIV proteins except gag and pol [56–59]; 5) EcoHIV mouse generated by inoculation of mice with a chimeric HIV that uses species-specific cellular receptors to enter mouse cells [60, 61]; and 6) humanized mice generated by transplanting human cells or tissues into genetically modified immunodeficient mice [62]. Further, the feline model of HIV is the only spontaneously occurring model of immunodeficiency [63–66], and the infected felines elicit high acute viremia, depleted CD4+ and CD8+ cells, and the establishment of chronic infection, whereby the virus is controlled naturally [67]. The target cells of FIV and HIV/SIV are distinct, and with the structure of the reservoir being different, the use of the FIV model for HIV and drug abuse research is likely limited [68]. While outcomes from a substantial number of laboratory-based findings reveal that drug abuse could exacerbate HIV-1 disease progression, epidemiological studies have shown mixed results. This review provides an overview of various animal models used in this field and a careful evaluation of methodological strengths and limitations inherent in the existing animal models to understand better how these animal models complement each other.

2. HIV and drug abuse comorbidity

Drug abuse is a prominent comorbidity and a contributor of HIV-1 infection. It is common among HIV-1-infected individuals and about 50% of people living with HIV-1 report current or past drug abuse history. HIV-1 infection and drug use intersect epidemiologically, and their combination can result in complex biological effects on the brain and alterations in behavior [69–72]. The extent to which drugs affect persons’ health with HIV-1 depends on many factors, including drug characteristics, use patterns, stage of HIV disease and its treatment, comorbid factors, and age [69, 73, 74]. HIV-infected patients who inject drugs have exhibited delayed HIV diagnosis, insufficient HIV care, poor antiretroviral adherence, treatment outcomes, and accelerated disease progression [69–71, 74–78]. Behavioral disorders, including depression, have also been identified as common critical comorbidities in patients with substance use disorders and/or HIV disease [79–82]. The literature survey on clinical studies shows that the prevalence of depression is high among HIV-1 infected people with substance abuse [69, 83–87]. HIV-mediated immune response and chronic neuroinflammation have been reported as significant factors underlying neurocognitive disorders, a severe complication of HIV-1 infection [3]. Previous reports have demonstrated that drug abuse contributes to increased viral load in the brain via increasing the infiltration of HIV-infected cells and accelerated glial infection, in turn, contributing to impaired immune responses and increased release of inflammatory mediators as well as neurotoxins [25, 88–93]. Several studies have also reported infiltration of HIV-infected cells to the brain and the role of infected glia as contributors of neuronal dysfunction and modulation of brain circuits, leading to accelerated drug abuse and addiction [93–98]. The development of models that accurately recapitulate HIV-1 infection and disease progression, and the comorbidity of drug abuse patterns in humans, could help understand the effects of drugs of abuse on HIV-1 infection and the development of effective therapeutics and vaccines. Also, a growing body of data links genetics to the risk of addiction [99, 100]. For example, the role of Caenorhabditis elegans as a model to study the molecular and genetic mechanisms of drug addiction has been elegantly reviewed [101]. The potential role of Drosophila melanogaster for the identification of target genes for drug addiction has also been reported [102]. Rodent models are also instrumental in establishing a link between genes and addiction [102]. This review discusses the several experimental models used to study the neurological effects of drugs of abuse on HIV-1 infection and their limitations.

3. HIV-1 Tg26 transgenic mouse model

HIV-1 Tg26 Transgenic mice harboring a non-infectious HIV-1 transgene pNL4-3: d1443 encoding all the HIV-1 genes except gag and pol were developed by Dr. Abner L. Notkins group in 1991 [103]. pNL4-3 (HIV-1 NL4-3 Infectious Molecular Clone) is a 7.4-kb recombinant proviral clone consisting of DNA from HIV-1 NY5, and LAV is used for in vitro manipulations of the HIV-1 proviral sequences [104]. The HIV-1 transgene pNL4-3: d1443 was generated by deleting the 3.1 kb sequence between the Sphl site at base position 1443 and the Bal1 site at position 4551 in pNL4-3 [105]. As shown in Fig. 1, the construct lacking the gag/pol region after deleting the 3.1-kb sequence was microinjected into fertilized mouse eggs derived from Friend leukemia virus B (FVB/N) mating [103]. The Tg26 mice are normal in appearance and weight and express high levels of HIV-1 transcripts including gp120, Tat, Nef, Rev, Vpu, Vpr, and Vif in the skin, skeletal muscle, and low levels in brain, lung, intestine, kidney, spleen, and thymus [106]. The homozygous mice show diffuse scaling of the skin and psoriasis-like lesions shortly after birth and develop progressive HIV-associated nephropathy (HIVAN)-like renal disease, comparable to that observed in human AIDS patients, and these mice do not survive beyond 3-6 weeks of age [106, 107]. The heterozygous Tg26 mice, on the other hand, have a much longer lifespan and develop the renal disease between 60 and 250 days [107, 108]. Thus, this model was widely used in studying disease pathogenesis, genetic susceptibility, and therapeutic intervention of HIVAN owing to its clinical relevance to ART-treated HIV-1-infected patients and resemblance to human HIVAN [107, 109].

Figure1.

Schematic of generating the pNL4-3:d1443 proviral transgene

Since most of the Tg26 mice generated on the FVB/N background had short lifespans and developed kidney disease, it posed a limitation in studying other HIV-associated disorders. Thus, HIV-1 Tg26 transgenic mice backcrossed onto other genetic backgrounds were studied, including CBA, DBA/2, CAST/Ei, C3H/He, BALB/c, DBF1, and C57BL/6. Among these mice strains, Tg26 mice backcrossed with C57BL/6 showed either normal renal histology or only mild mesangial expansion without overt podocyte damage [110, 111] and became a reliable model for the study of HAND as these mice have longer life expectancies. The viral mRNAs encoding Tat, gp120, vpu, and nef were detected in the brains of Tg26 mice, especially in the SVZ and SGZ neurogenic regions [112]. Moreover, the HIV protein Tat was detected in the homogenates from brain microvessels of Tg26 mice and was found to have higher expression levels in the older than younger Tg26 mice [113].

Interestingly, the expression of viral transcripts in the brain was relatively low but was constitutively persistent, thus implicating its clinical relevance to chronic HIV-1 infection [112]. Furthermore, an earlier study that examined archival pediatric brain tissues from series of HIV-1 infected infants and children with severe NeuroAIDS found HIV-1 RNA transcripts localized to the neurogenic zones, which was in agreement with the presence of HIV-1 viral transcripts in neurogenic zones of Tg26 mice [114]. Putatunda et al. demonstrated that Tg26 mouse neural stem cells (NSCs) had reduced neuronal but increased astrocytic differentiation. There were significantly higher amounts of quiescent NSCs, as well as significantly lower levels of active NSCs, proliferating neural progenitor cells, and neuroblasts in the subgranular zones of Tg26 mice. Also, newborn mature granule neurons in the dentate gyri of Tg26 mice had deficiencies in dendritic arborization, dendritic length, and dendritic spine density [112]. It was also found that Tg26(+/−) mice exhibited no significant differences in behavioral functioning, contextual fear conditioning, or cued fear conditioning responses as in wild-type mice. Male Tg26(+/−) mice, on the other hand, revealed significantly impaired short and long-term spatial memory, while female Tg26(+/−) mice had impaired spatial learning abilities and short-term spatial memory, which related to fewer quiescent neural stem cells and neuroblasts in the hippocampi compared with wildtype mice [115]. Thus, these studies indicated that the Tg26 mice model could also help understand the mechanisms underlying HAND pathology and study sex-related differences associated with the disease.

HIV Tg26 mice have also been used to study the interaction of HIV-1 infection and drug abuse [72, 116]. For example, studies using HIV-1 Tg26 mice demonstrated that morphine inhibited the growth of intestinal organoids derived from Tg26 mice by inhibiting the Notch signaling pathway [117]. Although several studies reported using the Tg26 mice [62, 63, 31], this model is not without limitation since the extent of translation of the viral transcripts is unknown.

4. HIV Nef mouse model

The accessory Nef protein is expressed by the HIV genome and is one of the six accessory proteins involved in T-cell activation and the establishment of persisting infection. The persistence of HIV infection is primarily based on the ability of Nef to downmodulate the surface levels of essential molecules at the immune synapse, i.e., major histocompatibility complex-I (MHC I) and (MHC II) on antigen-presenting cells, CD4 and CD28 on present on T helper cells [118]. Several models of HIV Nef have been established to date. Briefly, transgenic mice expressing Nef in the T-cells and other regulatory elements - CD38, CD2, and TCR β have been developed [119, 120]. While these models exhibit loss of T-cells and alterations in their activation accompanied with pathological lesions in the heart and kidney, none of these models exhibit multiorgan syndrome or a disease phenotype similar to AIDS. Expression of Nef in the double Tg mice (TRE/HIV nef × CD4C/rtTA) was also induced by doxycycline administration and resulted in the aging phenotype [121, 122]. Another type of Nef transgenic mice model with the expression of Nef in the microglia was developed. These mice demonstrated hyperactive behavior in male mice, with the production of CCL-2 in the striatum, and exhibited a dysfunctional dopaminergic system, thereby underscoring their utility as a model for HAND pathogenesis [123]. There are, however, several limitations in this model - since this mouse model only expresses Nef, it precludes the contribution of various other viral proteins [124].

Additionally, humanized mouse models have also been used for the study of HIV Nef. In this model, wild type or the Nef-defective HIV-1_JR-CSF/HIV-1_NL4-3 were injected into the thymic organoids grown from human fetal thymus and liver implanted under the kidney capsule of SCID mice. Both the Nef defective viruses showed attenuated growth properties relative to wild type. Nef defective HIV-1_NL4-3 lost the ability to prevent thymocyte loss, while active viral infection in wild-type virus caused loss of thymocytes within six weeks of infection [125]. These results demonstrated that HIV-1 Nef is required for in vivo viral replication in the thymus. Another humanized mouse model for HIV-1 Nef (LAI, CXCR4 tropic) showed significant depletion of CD4+ T cells. These results collectively demonstrate that viral Nef contributes to T-cell loss during HIV-1 infection. Infection of this mice model with HIV-1 JRCSF (CCR5 tropic) strain resulted in the loss of CD4+T cells in various organs but not in the peripheral blood [124]. Results from these studies implicated the role of Nef in HIV viral replication and CD4+ T cells. Studies also showed the role of Nef in HAND pathogenesis-transgenic mice in neuropathology, e.g., vacuolar myelopathy and changes in the oligodendrocyte phenotype in the brain [126, 127]. In another study, it was shown that rats implanted unilaterally with astrocytes expressing Nef, exhibited impairment of novel location and novel object recognition compared with controls that were implanted with astrocytes expressing the green fluorescent protein. Additionally, this impairment correlated with an increased expression of the chemokine ligand 2 (CCL2) in the implanted Nef transfected astrocytes followed by the infiltration of peripheral macrophages into the hippocampus at the site of injection. These rats also showed bilateral loss of CA3 neurons. This study thus showed that continued expression of Nef by astrocytes in the absence of viral replication has the potential to contribute to HAND [128]. Interestingly, cell culture studies have demonstrated that Nef overexpressing astrocytes (via adenovirus vectors) can release Nef in the extracellular vesicles, which, in turn, can be up taken by the neurons, leading to neuronal dysfunction in HAND [129]. Although few studies indicate the role of Nef in HAND, further detailed studies are required to dissect the role of Nef in other in vivo models of HAND.

5. Doxycycline-inducible HIV-1 Tat transgenic (iTat) mice model

The doxycycline-inducible HIV-1 Tat transgenic (iTat) mice model was generated from crossbreeding two separate transgenic lines -GFAP-rtTA mice (G-tg) and TRE-Tat86 mice (T-Tg). The G-tg transgenic line was obtained using a Teton-GFAP DNA fragment released by Xhol and Pvull digestion of the pTeton-GFAP plasmid, while the T-tg transgenic line was obtained using a DNA fragment (1189 bp) released by the same restriction digestion (Xho I and Pvu II) of the pTRE-Tat86 plasmid. These two DNA fragments were separately microinjected into fertilized eggs of F1 females (obtained from a mating between C3HeB and FeJ mice), then crossed with C57BL/6 mice to generate stable transgenic lines [50]. GFAP-rtTA mice and TRE-Tat86 transgenic mice were crossbred to obtain iTat mice carrying both GFAP-rtTA and TRE-Tat86 transgenes. The Tat expression levels were regulated by both the astrocyte-specific glial fibrillary acidic protein (GFAP) promoter and a doxycycline (Dox)-inducible promoter in the iTat mice model [130]. Expression of Tat mRNA levels in various organs or tissues of iTat mice was assessed to ensure the designed constructs specifically expressed Tat within the brain [50, 131] (Figure 2). It was shown that systematic exposure to Dox-induced Tat expression only in the brain and not in other organs or tissues such as the eye, heart, kidney, liver, lung, spleen, or thymus [50]. The protein levels of Tat in iTat mice’s brains were found to range from 1-5 ng/ml, which agrees with the levels of Tat reported in the brains of HIV-infected individuals [51, 52]. The inducible Tat mice are thus a valuable tool for understating Tat neurotoxicity and its role in HIV-1 neuropathogenesis. This model significantly improves over direct injection of recombinant Tat protein into the brain since the injection causes secondary damage to the brain [52].

Figure 2.

Schematic of generating the iTat mice

The utility of this model is also evident by the fact that Tat expression in the brain of iTat mice also leads to several developmental and behavioral abnormalities, including failure to thrive, hunched posture, tremor, ataxia, slow cognitive and motor movement, seizures, and premature death [50], similar to some of the cognitive changes observed in humans with HAND. Extensive neuropathological changes have also been reported in the iTat mice, including astrocytosis, neuronal death, degeneration of neuronal dendrites, morphological changes of astrocytes and microglia, and infiltration in the CNS by monocytes and activated T lymphocytes [50, 51, 132, 133]. Furthermore, increased expression of cytokines and chemokines has also been reported in the brains of Dox exposed iTat mice, including monocyte chemoattractant protein-1 and monocyte chemoattractant protein-2 (MCP-1/MCP-2), macrophage inflammatory protein-1α and protein-1β (MlP-1α/-1β), RANTES, Interferon gamma-induced protein 10 (IP-10). Thus, this model encompasses several features that recapitulate HIV-induced neurocognitive disorder findings both before and in the era of combination antiretroviral therapy (cART) [52].

Over the past two decades, there has been an escalation in the number of studies using the iTat mice to examine the direct effects of viral Tat protein on astrocytes and the molecular mechanisms of Tat-induced GFAP expression/astrocytosis [52]. Tat neurotoxicity [51, 134]. Tat-mediated impairment of neurogenesis [135]. Tat-induced loss of neuronal integrity [135], and exosome-associated Tat release and uptake [136]. For example, it has been demonstrated that Tat affected neural progenitor cells (NPCs) and neurogenesis through Notch signaling in the iTat mice, which underscores the potential of developing Notch signaling inhibitors for NeuroHIV therapeutics [135]. Interestingly, many reports have also utilized the iTat model to address Tau processing [137], comorbidity of substance abuse [138–144], sex differences [145], and aging [146]. For example, the effects of long-term Tat expression in the brain on behavior, pathology, and epigenetic landscape were studied by feeding the iTat animals a Dox-containing diet for 12 months, which was suggested to simulate people infected with HIV-1 for ~5 decades. These findings demonstrated that long-term Tat expression in the brain led to poor memory and impaired motor functioning, brain region- and sex-dependent dysregulation of neuropathological marker expression, DNMT3B expression genomic DNA methylation, thereby underscoring the critical role of HIV Tat in accelerated aging [146].

While this model has been used extensively in the field, providing relevant information on neurobehavioral and neuropathogenesis of HAND, it also has several limitations. First and foremost, the iTat model is not an infectious model and, while it can be argued that it mimics the persistence of Tat in the brain (as is evident in infected individuals on cART), the limitation is that it is Tat-centric, ignoring the impact of other residual proteins such as gp120, gp140, Nef, Vpr, and Rev [147–149] in the CNS. Secondly, how does this model correlate to the HAND neuropathology in cART-treated individuals? Furthermore, while it has been reported that HIV-1 infects up to 19% of perivascular astrocytes in individuals with HAND [150], these cells are not the primary producers of viral Tat protein in the brain of HIV-infected patients.

6. EcoHIV mouse model

HIV-1 does not infect rodents owing to viral entry restriction. To overcome the viral entry restriction, Potash et al. designed a chimeric virus that contains most of the HIV genes, and that can infect mouse immune cells by replacing the HIV gp120 gene with the ecotropic mouse leukemia virus gp80 gene, thus enabling the entry of the virus into mouse cells expressing the cationic amino acid transporter [151]. The resulting chimeric virus constructs, EcoHIV, productively infects murine lymphocytes but not human lymphocytes in culture. Adult immunocompetent mice are readily susceptible to infection by a single inoculation of EcoHIV, as evidenced by the virus detection in splenic lymphocytes, peritoneal macrophages, and the brain [151].

To date, two chimeric viruses have been utilized in studies investigating gene expression in both the periphery and CNS. Specifically, the EcoHIV chimeric virus is built on a backbone of clade B NL4-3 and EcoNDK, built on a backbone of clade D HIV-1. Both EcoHIV and EcoNDK have been shown to establish a systemic infection in mice after one inoculation [151]. Mice infected with these viruses reproduce the characteristics of HIV-1 infection in humans, such as viral infection of immune cells, i.e., lymphocytes and macrophages, induction of antiviral immune responses, neuroinvasion, and expression of inflammatory and antiviral factors in the brain. Infection of mice with EcoNDK also increases proinflammatory factors such as interleukin-1 β and STAT-1 in the brain [151]. A recent study has shown that EcoHIV infection of mice can also establish latent viral reservoirs in T cells and concurrently show active viral replication in macrophages, leading to the induction of neurocognitive impairment [61].

The EcoHIV model has been valuable for many applications, including antiviral drug screening [60], development of vaccines [152, 153], studying the effects of opioids on early HIV pathogenesis in the gut [154], investigating the therapeutic potential of biologicals such as insulin for HAND [155], and more recently, for HIV viral latency, pathogenesis and neurocognitive impairment [61, 155–157]. Valuable insights have been gleaned from the EcoHIV rodent model. Firstly, this model serves as an infectious model of HIV-1, resembling HIV infection in the era of cART [158, 159]. This model has provided utility in the preclinical evaluation of anti-retroviral drugs and has been found to provide rapid, statistically robust, and inexpensive evaluation of antiretroviral drugs in vivo [60]. Similarly, it has also proven useful to test novel vaccine candidates [152, 153]. Biologicals such as insulin were also tested in vivo via intranasal delivery for their protective effects against HAND using the EcoHIV-infected mice [155]. The EcoHIV model has also been reported to be a valuable model for studying the effects of opioids in HIV-induced pathology during the early stages of infection in the gut [154].

The EcoHIV model has also played an essential role in understanding how methamphetamine potentiates HIV infectivity in NPCs [160]. In this study, the authors infected mouse NPCs with EcoHIV in the presence or absence of methamphetamine (50 or 100 μm). They found that methamphetamine increased HIV infectivity of NPCs, involving the NFκB/SP1-dependent activation of HIV LTR, thereby suggesting that methamphetamine exacerbates neurocognitive dysfunction [160]. The EcoHIV model, similar to other models, is not without limitations. The chimeric virus infects multiple cell types, which may not adequately reflect the human scenario. Additionally, viral levels and the spread of infection in this model do not precisely mimic the humans infected with HIV-1.

7. HIV envelope glycoprotein 120 (gp120) Transgenic (Tg) mouse model:

HIV envelope glycoprotein 120 (gp120) is one of the most prominent viral antigens produced in HIV-infected cells and is shed into the extracellular space as a soluble protein, which, in turn, has the potential to interact with uninfected cells [161]. The toxicity associated with gp120 was examined as a model to explore the neuropathological and neuropsychiatric manifestations associated with AIDS [147, 162–166]. The HIV gp120-transgenic (Tg) mouse model was generated by inserting the HIV env gene that encodes soluble gp120 under the control of the GFAP promoter, such that it is expressed in astrocytes [167]. The transgene is highly expressed in the brains, specifically in the neocortex, olfactory bulb, hippocampus, tectum, selected white matter tracts, and other areas associated with astrocytes. The neuro pathophysiology, including decreased synaptic and dendritic density, frank neuronal loss, increased numbers of activated microglia, and pronounced astrocytosis and behavioral phenotypes found in human HIV patients, has also been found in this Tg mouse model [166, 167]. An interesting study using this model demonstrated that gp120 expression was a useful indicator for assessing the pathophysiological effects on neuronal networks and showed divergent changes in short- and long-term potentiation in the CA1 hippocampus [168]. Although this Tg model only expresses viral protein gp120, this model has utility for the study of HAND pathology and recapitulates human HIV patient brains on cART therapy and low levels of replicating virus [166]. Many studies have shown that these mice’s brains mimic the pathological findings observed in HAND patients, including differentially regulated genes and cellular signaling pathways, activation of innate immunity, impaired neurogenesis and cognition, and learning and motor deficits [53, 158, 166]. These animals also exhibited impaired proliferation and differentiation of NPCs [169, 170]. Interestingly, gp120 Tg mice also demonstrate age-dependent impairment in open field activity and spatial reference memory, reflecting cognitive and neuromotor deficits seen in HIV-1-infected individuals [165]. Another study using the gp120 Tg mouse model demonstrated that pharmacological abrogation of CCR5 signaling exerted a novel protective effect on lipocalin-2 associated neuronal injury and behavioral impairment [53]. In another study, it was shown that gp120 Tg mice exhibited DNML1-associated downregulation of mitochondrial fission (Dynamin-1-like protein-associated mitochondrial fission) in neurons to that observed in the brains of HAND patients [171]. A recent study reported that gp120-tg mice showed NLRP3- dependent pyroptosis and IL-1β production in microglia [147]. Importantly, inhibition of NLRP3 inflammasome activation in microglia via chronic administration of MCC950, a novel selective NLRP3 inhibitor, to gp120 Tg mice attenuated neuroinflammation and neuronal death, while restoring neurocognitive functioning [147].

Drugs of abuse and HIV-1 are implicated in the progression of neurodegenerative effects such as astrogliosis [172], compromised blood-brain barrier (BBB) integrity, and excitotoxicity in the brain, as well as the progression of neurocognitive decline [173–176]. An earlier study using the gp120 Tg mouse model showed an increased preference for methamphetamine and increased sensitivity to methamphetamine-induced conditioned reward [173]. While this model has provided value in the field, the limitation of this model is that constitutive expression of the Tg throughout the lifespan of the animals can lead to early developmental changes before the onset of the desired pathological phenotype. Like the iTat mice, another important caveat of this model is that these animals only express one viral protein, which excludes the pathological effects of other viral proteins, thus failing to recapitulate the human clinical scenario of the disease [163, 166]. Crossbreeding of the Tg models expressing distinct viral proteins could help to elucidate the combined effects of various viral proteins in mediating neuropathogenesis.

8. Humanized mouse model

While the Tg mice models expressing select HIV-1 proteins (such as HIV-1 envelope and Tat) mimic aspects of HIV-1-related pathologies [108, 177–180], they only mirror specific pathophysiological effects of select HIV-1 proteins on the CNS [181] but do not recapitulate infectious virus. In light of this, the development of humanized mice models that can mimic the human immune system is of prime value. As stated above, all strains of mice are naturally refractory to HIV-1 infection. However, the humanized mice, which are immunodeficient reconstituted with human donor cells and tissues, offer a unique small animal model for studying HIV neuropathogenesis, prevention, and therapeutic development in vivo.

8.1. Severe combined immunodeficient (SCID)-HIVE model

The SCID-HIVE model was developed at Johns Hopkins University in the early 1990s [40] to investigate HIV-infected cells’ role in contributing to HIV-associated encephalitis (HIVE). In the original study, the SCID mice, which lack murine B and T cells, were intracerebrally injected with isolated human peripheral blood mononuclear cells (PBMC). Next, the PBMC-injected SCID mice and control mice were injected with HIV-1. As expected, the SCID mice injected with both PBMC and HIV-1 developed striking gliosis and pathology characteristic of HIVE. Subsequent studies of injecting HIV-infected human myeloid cells into SCID mice’s brains [39, 182–184] also resulted in pathologic changes similar to HIVE, such as encephalitis, astrogliosis, and an increased influx of mononuclear phagocytes, within a week. This model was also tested to assess cART treatmen’s effects in mitigating HIVE-associated pathologies [185–187]. The results demonstrated that cART could reverse HIV-mediated glial activation, mononuclear cell infiltration, and neuronal dysfunction, suggesting that replication of HIV-1 directly contributes to brain pathology in the SCID-HIVE model.

Recently, the SCID mouse model has been used to evaluate strategies for reversing HIV-induced neuroinflammation. Baricitinib, a JAK 1/2 inhibitor, which has been approved for rheumatoid arthritis, demonstrated promise in the inhibition of pro-inflammatory cytokines such as IL-6, D-dimer, CRP, TNF-α, IFN-α/β [188]. In the SCID mice, baricitinib treatment was shown to cross the BBB and reverse HIV-1-mediated behavioral abnormalities. Furthermore, baricitinib reduced glial activation, as evidenced by decreased expression of MHCII/CD45 (microglial activation) and GFAP (astrogliosis). Additionally, baricitinib treatment also significantly decreased p24+ human macrophages in mouse brains [188]. CEP-1347, an inhibitor of mixed lineage kinases, was also shown to reduce microglial inflammation as measured by decreases in Iba-1 expression [189] in the SCID-HIVE mouse model. Also, SCID-HIVE mice treated with B18R, a novel IFN-α inhibitor that crosses the BBB, demonstrated decreased numbers of mouse mononuclear phagocytes that were accompanied by significant restoration of neuronal arborization, as measured by decreases in the expression of mouse microtubule-associated protein-2 (MAP-2), compared to untreated SCID-HIVE mice [190]. SCID-HIVE mice were also utilized as a model to address the comorbidity of drugs [191–194] and autoimmune diseases [195, 196].

One of the significant limitations of the SCID-HIVE model is that the direct injection of human cells into the brain results in a traumatic injury that manifests as a cellular activation in the brain. The pathology is more pronounced in mice injected with HIV-infected cells than in mice administered uninfected cells. Additionally, injection of the mature immune cells could also lead to graft-versus-host-disease, which results in a shorter lifespan of SCID-HIVE mice (four weeks or less). Thus, this limitation eliminates the use of the SCID-HIVE model for long-term analyses of HIV persistence [197].

8.2. CD34-transplanted NSG mouse model or hu-NSG

The NOD.Cg-Prkdc^scidIl2rg^tm1Wjl/SzJ (NSG, NOD SCID gamma) mouse was generated by introducing an Il2rg knockout allele into NOD SCID mice, producing a mouse strain that lacks functional B cell, T cells, and natural killer (NK) cells. In contrast to the NOD SCID mouse, NSG mice lack functional NK cells that prevent them from the development of thymic lymphoma (the primary cause of death of NOD SCID mouse), resulting in an increased life span (NSG = 16 months, NOD SCID = 8 to 10 months) [198], thereby making the NSG mouse a superior model for long-term HIV-1 study.

The hu-NSG model is derived from NGS mice engrafted with human CD34+ hematopoietic stem cells, isolated from the cord blood, bone marrow, or fetal liver either by intravenous or intrahepatic injection [62, 199, 200]. HIV-1-infected NSG mice were found to exhibit systemic HIV-1-infection and CNS viral burden, leading to the development of HIV-associated pathologies that are accompanied by depletion of CD4+ T-cells, viral replication, CNS infiltration of HIV-1 infected monocytes and macrophages, resident microglial and astrocytic activation, neuroinflammation and cognitive abnormalities [62, 181, 201]. This hu-NSG model has been successfully utilized to evaluate multiple HIV-1 therapeutic strategies in vivo. For example, Wu et al. investigated the prophylactic protection effect of an engineered broadly neutralizing antibody (bs-bnAb, BiIA-SG) in the hu-NSG model. Treatment with this antibody was found to eliminate infected cells in hu-NSG humanized mice. Additionally, the antibody treatment combined with cART exposure could delay the viral rebound in this hu-NSG humanized mouse model [202]. Kumar and colleagues successfully used a CD7-specific single-chain antibody conjugated to the oligo-9-arginine peptide (scFvCD7-9R) to achieve T cell-specific anti-CCR5 and antiviral silencing RNAs (siRNA) delivery in hu-NSG mice. This was found to control viral replication and prevent disease-associated CD4 T cell loss [203]. Recently, Sonic Hedgehog (Shh) mimetic, a Smoothened Agonist (SAG), was found to reduce CNS viral burden immediately after HIV-1 transmission, while also conferring extended neuroprotection via increasing BBB integrity, assessed by elevated levels of tight-junction protein, Claudin 5, and decreased levels of S100B in the periphery of HIV-1 infected huNSG mice [204].

8.3. Myeloid only mouse (MoM) model

Although several studies implicate that monocytes and macrophages are the critical cells infected by HIV-1, both in vitro and in vivo [205–207], the question of whether myeloid lineage cells serve as HIV-1 reservoirs in the context of cART treatment remains an area of high investigation. Recently, Honeycutt et al. generated a myeloid-only mouse (MoM) by injecting CD34+ human hematopoietic stem cells, reconstituted with human myeloid and B cells in the absence of human T cells, into pre-conditioned NOD-SCID mice [208]. A major benefit of MoM is that it rules out the possibility that macrophages could become HIV-1 positive via the phagocytosis of infected CD4+ T cells [209–211]. Through the use of this Mom mouse model, Honeycutt et al. demonstrated that macrophages could be infected by HIV-1, independent of T cells, to support efficient HIV-1 replication in vivo [212]. Additionally, the withdrawal of suppressive cART from MoM resulted in viral rebound after seven weeks [208], thus supporting the hypothesis that long-lived tissue macrophages could serve as HIV-1 reservoirs [208, 213].

8.4. Humanized microglial mouse

While most studies of HIV-1 latency focus on the peripheral population of T cells, the brain also contains a distinct reservoir of HIV-1 infected cells in the microglia. Human perivascular macrophages, microglia, and astrocyte are critical for simulating HIV-1 induced neuropathogenesis and the study of brain virus reservoirs. Immune deficient humanized mice are susceptible to HIV-1 infection and have been shown to serve as valuable tools for HIV-1 infection of the CNS. Low numbers of human cells and even lack of human microglia have been detected in humanized mice’s brains by histology and qPCR [212, 214, 215]. It is essential to generate the human microglial mice for targeting CNS HIV infection and for the study of HIV-1-mediated neuropathogenesis.

Recently, Mathews and colleagues developed a new humanized microglial mice model that comprises an immunodeficient strain supplemented with human interleukin-34 (IL-34) transgene to support microglial development [216]. Herein, the authors first generated the hIL-34 expressing transgenic NOG mice (NOG-hIL-34) by microinjecting fertilized eggs obtained by mating NOG (NOD.Cg-Prkdcscidil2gtmlSug/Jic) and NOD (NOD/ShiJcl) mice with a linearized DNA vector (pCMV6-XL4) containing human IL-34 cDNA under the control of a CMV promoter. Next, this NOG-hIL-34 mouse has engrafted intrahepatically with the umbilical cord blood-derived CD34+ hematopoietic stem progenitor cells (HSPCs) to generate the humanized microglial mice (CD34-NOG-hIL-34). HSPC-derived peripheral macrophages generated microglial-like cells in the CNS and expressed human microglial cell markers such CD14, CD68, CD163, CD11b, ITGB2, CX3CR1, CSFR1, TREM2, and P2RY12 in the CD34-NOG-hIL-34 humanized mouse model. Intriguingly, these mice expressed features resembling the pathobiology of HIV-1 infection in the CNS, including activation of astrocytes and microglia, a productive viral infection of microglia, synaptic alterations, and inflammatory responses, thus underscoring the role of this model in neural-glial cross talk and in studies aimed at targeting/eliminating the viral reservoirs in the brain. In another study, when iPSC-derived hematopoietic progenitor cells (HPCs)-primitive macrophage progenitors were transplanted into the brain ventricles and overlying cortices of the immune-deficient, humanized CSF1 mouse pups, it was shown that the expression of human CSF1 was required for long-term engraftment of human microglia in the murine brain [217]. Furthermore, it was also shown that the microglia that developed from the transplanted HPCs had transcriptional profiles similar to those of microglia from the human brains. These findings thus suggested that human iPSC-derived microglial chimeric models can mimic human microglia development and its impact on neuronal development, synaptic plasticity, and behavioral performance of the animals [217, 218]. There have also been reports wherein embryonic stem cells (ESCs)-derived human microglia were used to generate microglia humanized mouse model. These authors differentiated H9 ESCs into microglia by using IL-34, colony-stimulating factor 1 (CSF1), transforming growth factor-β (TGF-β) and CX3CL1, followed by transplanting cells into the brains of Rag2−/− Il2ry−/− hCSF1KI mice (hCSF1KI) at postnatal day 4. To create a permissive environment for human microglia integration, the neonates were pretreated with CSF1 receptor (CSF1R) inhibitor BLZ945 [219] that eliminates an average of 53% ± 7% of host microglia. Eight weeks later, H9-microglia represented 9 ± 5% of the total microglial population with a mosaic distribution across multiple areas of the brain. The H9-microglia demonstrated a complex ramified morphology that could be detected by the contribution of microglia in neurological diseases.

9. Rat models: HIV-1 Transgenic rats

The HIV-1 transgenic (Tg) rat model (HIV-1 Tg) developed by Reid et at. expresses a noninfectious HIV-1 provirus Tg that has both the gag and pol genes deleted from the infectious clone of an integrated proviral plasmid (pNLS-3) [56]. The approach involved deleting an overlapping fragment containing 2 of the 9 viral genes, i.e., the gag gene at the 3′ region and the pol gene at the 5′ region, resulting in a non-infectious provirus pEVd1443 [56]. This non-replicative provirus is regulated by the viral long terminal repeat, which activates the transcription of the env gene plus the regulatory (tat, rev) and the accessory (vif, vpr, vpu, and nef) genes.

Here, the non-infectious provirus was microinjected into the fertilized eggs derived from a Fisher/NHsd (F344) rat and Sprague Dawley rats to produce a female Tg founder [56]. The Tg was incorporated into one copy of the two alleles of the HIV-1 Tg rat. When these rats are mated with a wild type inbred F344 rat, the offspring are either HIV-1 Tg or wild type. The HIV-1 Tg rats have an easily distinguishable phenotype that includes opaque cataracts, wasting, skin lesions, neurological signs, respiratory problems, and nephropathy [56, 220]. The wild-type littermates without cataracts or standard F344 rats are usually used as control animals.

The HIV-1 Tg rat model is suitable for studying the effects of HIV-1 viral proteins in both the periphery and the CNS. Consequently, the HIV-1 Tg rat model has been utilized in studies of the neurological effects of HIV-1 [221, 222], investigations focused on the interaction of HIV proteins with drugs of abuse [223], and behavioral studies [57]. Several impressive results have been obtained using this HIV-1 Tg rat model. The HIV-1 Tg rat mimics HIV-infected patients on cART, owing to low levels of viral proteins. This model has been used to study the long-term effects of these persistent viral proteins on disease pathogenesis. For example, one study using this model demonstrated astrocytic dysfunction/damage and dopaminergic neuronal loss/dysfunction compared with wild-type controls. Interestingly, both these parameters worsened in the striatum with the animals’ increasing age [222]. Another study on immune activation, viral gene expression, and neurotoxicity in HIV-1 Tg rats demonstrated that activated mononuclear phagocytes and astrocytes expressing the HIV gene products in the brain were associated with increased neurodegeneration [221].

The HIV-1 Tg rat model also provides an opportunity to study mechanisms underlying HIV-related neurological and cognitive dysfunction, a limited research area due to the lack of an adequate animal model [57]. In a study, Vigorito et al. evaluated the HIV-1 Tg rat model, using the Morris water maze for testing learning and memory deficits and found that these rats showed a deficit in learning how to swim to the location of the hidden platform, thus suggesting the validity of HIV-1 Tg rats as a model for behavioral processes underlying HIV-related neurological dysfunction [57]. In the context of drug abuse, this model was used to demonstrate that methamphetamine self-administration increased compulsivity, inflammation, and neural injury in these animals compared to the wild-type controls [223, 224]. The HIV-1 Tg rat model was also used in behavioral studies with drug abuse [225–227]. This model also is fraught with limitations. As described earlier, this is a non-infectious model and, thus, is not suitable for studies investigating viral progression or replication or for studying the impact of novel antiviral drugs on viral replication [228].

10. Rodents with direct injection of HIV proteins

Several research groups have used the direct injection of HIV proteins into the brain to study the effects of individual proteins such as Tat or gp120 in modulating cellular processes involved in neurotoxicity and neurodegeneration. Among some of the earliest studies, Hayman et al. injected synthetic peptide analogs derived from the basic region of HIV-1 Tat protein into the rat brain and found Tat to be a potent neurotoxin [45]. This was confirmed by another study that also demonstrated that blocking TNF-alpha with pentoxifylline following Tat injection led to decreased IL-1β and iNOS expression that was accompanied by a reduction in the volume of pathological lesions, indicating that the Tat-induced lesions could be likely mediated by TNF-α production [229]. To study the neurotoxicity of HIV-1 proteins gp120 and Tat in the rat striatum, Bansal et al. stereotactically injected recombinant proteins (gp120, Tat, or saline as a negative control) in the striatum of adult male rats. They also evaluated the area of tissue loss and the number of GFAP immunoreactive cells 7 days post injections [230]. These authors found that doses of gp120 (250 ng/μL or higher) and Tat (5 microg/μL or higher) produced a significant area of tissue loss and significantly increased numbers of GFAP reactive cells thus suggesting the involvement of gp120 and/or Tat in striatal toxicity. In another study, the injection of HIV-1 Tat protein into the rat striatum revealed that Tat caused an increase in protein carbonyl formation. This type of oxidative modification of proteins occurred early after Tat injection preceded Tat-mediated astrogliosis [44]. These results were part of the early evidence suggesting that increased protein oxidation could be an essential mechanism underlying Tat neurotoxicity [44]. It was also shown that intraventricular injection of HIV-1 Tat protein in rats caused a cascade of events leading to the influx of inflammatory cells, glial cell activation, and neurotoxicity [46].

The interaction of Tat with methamphetamine was studied following injection of Tat into the striatum and methamphetamine by i.p. injection [231]. These studies showed that Tat injection, with concurrent administration of methamphetamine, profoundly increased chemokine and cytokine levels (MCP-1 and interleukin-1 alpha) in the striatum resulting in enhanced dopaminergic system damage [231]. The same group also utilized the Tat injection model to demonstrate that methamphetamine and Tat synergize to destroy dopaminergic terminals in the rat striatum [232].

Direct injection of HIV proteins into the CNS, as described above, demonstrates the utility of this model in elucidating the mechanisms involved in viral protein neurotoxicity and neuroinflammation. Results from studies using this model suggest specific pathways that could be blocked to attenuate viral protein-induced neurotoxicity and neuroinflammation. However, the limitation of this model is the inherent disadvantage of short-lived neuroinflammation in the CNS due to the rapid clearance of the injected proteins. Another caveat of this model is that stereotaxic injection of the viral proteins can result in a traumatic brain injury that can confound interpretation of results because it is challenging to differentiate injection-induced injury from infection-induced injury [158].

11. Simian Immunodeficiency Virus (SIV)

11.1. Natural SIV hosts

Around 40 different African primates species are natural hosts for SIV and can be infected with their own specific strains of SIV. However, these naturally occurring SIV infections do not result in disease due to thousands of years of virus-host co-evolution and are thus not useful as pathogenic infection models [233, 234]. Among the different natural hosts, Sooty mangabey (Cercocebus atys) carrying SIVsmm virus and African green monkey (Chlorocebus aethiops) carrying SIVagm infection have been studied due to their availability at primate centers in the United States and Europe [234]. The Sooty mangabey model is of much interest as the transmission of SIVsmm to macaques resulted in the development of SIVmac, while transmission of this virus to humans gave rise to HIV-2 [234]. The non-human primate models of SIV have many advantages, including vaccine research, and as SIV is genetically and morphologically close to HIV, this model also useful to study the effects of drug abuse [235–239].

11.2. Non-natural SIV hosts

In contrast to the fact that the natural hosts of SIV do not exhibit chronic infection, Asian macaques are more vulnerable to infection with certain SIV strains and have thus been the most widely accepted models of HIV-1 infection. These models have been used extensively in studies to characterize disease pathogenesis and the development of vaccine development. The most commonly used non-human primate models of AIDS comprise the Indian rhesus macaques (Macaca mulatta), pig-tailed macaques (Macaca nemestrina), and the cynomolgus macaques (Macaca fascicularis) [240]. These macaque models have a unique evolutionary history that significantly impacts the outcome of SIV infection [240]. Indian origin Rhesus macaques are the best characterized and likely most utilized non-human primate model for the study of AIDS. The most commonly used virus strains for these macaques are SIVmac251 and SIVmac239, which cause productive infection with high viral loads and progressive loss of CD4+ T cells, particularly in the GALT [241]. Both of the viruses are neurotropic and have been well studied in models of NeuroAIDS [242–244], as discussed later. Tripartite-motif-containing protein 5α (TRIM5α) was previously known to inhibit the post-entry block to HIV-1 infection of rhesus macaque cells [245]. Rhesus macaques encode highly polymorphic TRIM5 proteins [246]. Reports have shown that viral loads in SIVsmE543-3-infected macaques are 100- to 1,000-fold lower in animals expressing specific TRIM5α variants (with 339-TFP-341 in the SPRY (SPla and the Ryanodine Receptor) domain) than in animals expressing variants (with Q339 in the SPRY domain) [247]. The susceptibility of SIVsmE543-3 to certain rhesus macaque TRIM5α proteins likely reflects the incomplete adaptation of this virus to this host, mainly because SIVsmE543-3 was cloned after sequential passage in only two rhesus macaques [247]. These observations are essential, as SIVsmE543-3 and the closely related SIVsmE660 strain are more often used as a heterologous-challenge virus for rhesus macaques immunized with SIVmac239-derived antigens.

In addition to the Indian, Chinese, and Burmese rhesus macaques have also been used to study AIDS, albeit less frequently than the Indian rhesus macaques due to limited availability. SIVmac239, SIVmac251, and SHIV89.6P have been less pathogenic in Chinese and Burmese macaques than their effects in Indian rhesus macaques [248–251]. The difference in pathogenicity is reflected in viral loads, higher CD4+ T cell counts, and prolonged survivability of Chinese macaques. The underlying reasoning for these differences is unclear but can likely be attributed to immunogenetic differences [252, 253]. Some studies have shown the relation between inflammation and viral infectivity in these Rhesus macaques. In a study, Sandler et al. [254] utilized SIV-infected Indian rhesus macaques to investigate the role of inflammation in HIV infection, particularly the impact of type-I interferons (IFN-I). These authors found that the timing of IFN-induced innate responses in acute SIV infection significantly affected the overall disease course [254]. Another group of investigators showed that increased proliferation and infection of CD4(+) T(SCM) could contribute to SIV infection’s pathogenesis in Indian rhesus macaques. The dynamics of germinal center (GC) formation in lymphoid tissues following acute SIV infection was a significant predictor of disease in rhesus macaques. Hong et al. showed that local production of IL-21 in lymphoid tissue GCs was an indicator of better control of viral replication and slower disease progression [255]. Thus, limiting dysregulated lymphoid function may be a critical target for new therapeutics. In one study, it was observed that there were polymorphisms of the TRIM5α B30.2/SPRY domain in SIVsmm, which regulated the viremia in rhesus macaques. Viremia in macaques homozygous for the non-restrictive TRIM5α allele TRIM5Q was significantly higher than in macaques expressing two restrictive TRIM5alpha alleles TRIM5^TFP/TFP or TRIM5^Cyp/TFP. This study further showed that mutation of TRIM5α in SIVsmm (substitution of the amino acid (P37S and R98S) in the capsid region) resulted in the cross-species transmission of SIV in primates [256]. In fact, it has been observed that SIVsm804E is capable of establishing infection in the CNS at an early stage of infection and causes neuropathology in infected rhesus macaques at a high frequency (83%) using a single inoculum when animals with restrictive MHC-I or TRIM5α genotypes are excluded, which depicts that macaques with restrictive TRIM5α, are resistant to developing NeuroAIDS [257].

In addition to the Indian rhesus macaques, the pig-tailed macaques are the most used nonhuman primates for AIDS research. Pig-tailed macaques infected with SIVmac239 develop AIDS within 42 weeks of infection, compared to 70 weeks for rhesus macaques [258, 259]. In these macaques’ strains, the MHC genes are not as well characterized as in the rhesus macaques. However, several SIV epitopes that are recognized by CD8+ T cells, and the MHC class I molecules that present these epitopes, have been defined, such as Mane-A1*084, which inhibits an immunodominant epitope of the SIV Gag protein and is associated with decreased viral load following SIV infection [260, 261]. A unique feature of pig-tailed macaques is that they do not express TRIM5α; they encode the TRIMCyp isoform, which does not restrict HIV-1 [262–266]. The absence of TRIM5α in these macaques could explain the reports showing that SHIV (HIV-1 vif gene was replaced by the vif genes from SIVmac239 and HIV-2ROD) cause a low-level transient infection in pig-tailed macaques. Several studies on SIV infection in macaques have shown virus replication in macrophages in the brain leading to encephalopathy, which was correlated with the ability of the virus (SIV/17E-Fr [267]. SIVsmmFGb [268] to replicate productively in cultured macrophages derived from peripheral blood of animals.

In AIDS research, Cynomolgus macaques have been utilized to a lesser extent than have rhesus and pig-tailed macaques. This is primarily due to the low level of immunogenic characterization of this species. In terms of viral loads and CD4+ T cell turnover, investigations have revealed that SIVmac251 and SHIV89.6P are much less pathogenic in cynomolgus macaques than in Indian or Chinese rhesus macaques [250]. Interestingly, it was found that the TRIM5 genes of cynomolgus macaques are also quite diverse, encoding polymorphisms in both TRIM5α and TRIMCyp isoforms, including TRIMCyp proteins, that could differentially restrict HIV-1 [269]. The limited MHC diversity of these macaques has dramatically increased interest in using the cynomolgus macaques in AIDS research [270, 271]. These animals originated from a small population brought to the Mauritius island approximately 400–500 years ago by European sailors. To date, only seven MHC haplotypes have been identified in Mauritian cynomolgus macaques [272]. Thus, unlike other primate models, for which it is only possible to select MHC-defined animals, studies can be done in Mauritian cynomolgus macaques using haplotype-matched or completely MHC-identical animals. This unique feature was used to demonstrate an MHC heterozygote advantage for the ability to control SIV infection [273].

11.3. Non-human primate models of NeuroAIDS

Not all SIV strains have the same pathogenic effects on the CNS. Several SIV models have been studied to understand the pathogenesis of disease in the brain. The early models of NeuroAIDS utilized Indian rhesus or pigtailed macaques infected with different viral clones that resulted in immune suppression, macrophage tropism for CNS infection leading to AIDS, and SIV encephalitis (SIVE), in approximately 30 percent of rhesus macaques, and higher frequencies in pigtailed macaques [274]. Strategies involving depletion of CD4+ and CD8+ T lymphocytes and/or serial passage of virus through monkeys for enhanced macrophage replication and neurotropism have been used in different studies [257, 275–277].

A widespread and widely used rhesus model includes monoclonal antibody (mAb) depletion of CD8+ lymphocytes and infection with the viral swarm SIVmac251. This resulted in persistent and high viremia leading to the rapid onset of AIDS (3–4 months versus 2–3 years in non-lymphocyte depleted animals) and a high incidence of SIVE (greater than 80% versus approximately 30% in non-depleted animals) [278]. This model yielded several significant findings, including high monocyte turnover from the bone marrow [279], accumulation of macrophages in the CNS [280], recruitment of inflammatory macrophages into the CNS [281], elevated levels of soluble CD163 in plasma that correlated with neurocognitive impairment [282], accumulation of cardiac macrophages [283], and identified the role of macrophages in SIV neuropathogenesis [282].

Infection of rhesus macaques with both SIVmac251 and SIVmac239 leads to the generation of virus-infected cells in the CSF throughout the acute phase, which can last for 2 to 10 weeks [284–286]. During this acute phase of SIV infection, neuropathogenesis is characterized by meningitis and accumulations of mononuclear cells in the Virchow-Robin spaces in the brain [287]. Studies have shown that lesions caused by SIVmac239 are less severe; instead, meningitis disappears at the end of the acute phase of infection [284, 287]. Encephalitic changes leading to dense infiltrations of mononuclear cells lead to neuroinflammatory milieu [288]. Unlike human syndrome, inflammatory lesions in the meninges and neuropil occur even when the CD4+ T lymphocytes are not depleted [288]. Furthermore, SIV lesions are much more severe in the CNS compared with those seen in HIV-infected brains. Interesting studies have shown that SIVmac251-infected macaques have altered evoked potentials in the brain, a drastic increase in the number of activated CD8+ T cells and neuroinflammatory milieu in the brains during early SIV infection [289–292]. SIVmac182 infected macaques have also been shown to cause CNS neuropathogenesis. SIV-specific CD8+ CTLs were detected in the CSF of infected macaques early after infection [290]. This model has been used to study the role of immune responses in mediating CNS injury during the relatively early stage of SIV-induced disease. The primary advantage of these models is that they exhibit a very fast rate of AIDS progression post-SIV infection compared to decades it takes in humans infected with HIV-1. However, since SIVmac251 comprises of both macrophage-tropic and non-macrophage-tropic strains, it does not faithfully model native HIV transmission, which is essentially universally via non-macrophage-tropic transmitter/founder viruses. The molecular clone of SIVmac239, a non-macrophage-tropic transmitter/founder virus, is ideal for studying the mucosal transmission of HIV-1 but is not a good model for studying the CNS infection of HIV, where the virus predominantly targets the microglia and perivascular macrophages.

The other common model for SIV neuropathogenesis involves mAb depletion of CD4+ T lymphocytes before SIV infection, resulting in increased viral replication, progressively leading to AIDS [275]. Earlier depletion of CD4+T cells is associated with a higher percentage and incidence of infected macrophages in the periphery and the CNS. Additionally, this model also showed higher levels of SIV infection in the parenchymal microglia [275]. These data underscore the role of macrophages/microglia in AIDS pathogenesis. In addition to SIV, chimeric simian-human immune deficiency virus (SHIV) models have also been shown to exhibit rapid depletion of CD4+ T cells in Indian rhesus macaques, leading to the rapid development of AIDS and CNS pathology [293, 294]. These models demonstrate the significance of CD8+ T and CD4+ T lymphocytes in the development and progression of AIDS and NeuroAIDS. A third model of rapid and predictable progression to AIDS with a high incidence of SIVE involved pigtailed macaques infected with a combination of a neurovirulent molecular clone that replicates efficiently in macrophages, SIVmac/17E-Fr, and an immune-suppressive viral swarm SIVsm/Delta B670 that rapidly depletes CD4+ T lymphocytes within 2–4 weeks post intravenous inoculation [295, 296]. This model has been comprehensively used to study the mechanisms involved in CNS, peripheral neuropathy, and cardiac pathogenesis and understand innate immunity in the CNS and CNS-based therapeutics for humans [297–299]. The SIVmacR71/17E virus has been shown to be neurotropic and causes neuropathogenesis in Indian Rhesus macaques. For example, a study by Bokhari et al. has shown viral replication in the brain and histolytic nodules in the basal ganglia of this virus infected for 52 weeks [300]. Several studies have shown that there was neuroinflammation in the brains of the SIV infected macaques, and this neuroinflammation was due to activation of inflammasome [301–303] as well to deposition of neurotoxic amyloids in the brains [304]. SIV R71/17E infected monkeys also exhibit deficits in motor performance [305].

Several published studies have shown that serially passaged SIVSmE543 and SIVsm804E result in early viral entry into the CNS with a high incidence of SIVE. Gabuzda and collaborators similarly identified a variant of macrophage-tropic SIV env glycoprotein in the blood of SIV251-infected animals and developed a molecular clone that enters the CNS early after inoculation, resulting in a higher incidence of SIVE [257, 277]. Overall, all the models discussed above, with or without cART, provide fundamental data underscoring the role of viral evolution and macrophage tropism that facilitates rapid and consistent development of AIDS with neuropathogenesis, that takes into account the role of acquired and innate immunity in CNS disease. Studies with macaques have shown that monocyte trafficking in and out of the CNS establishes and maintains a CNS reservoir of productive SIV infection [282]. Earlier studies have shown that cART treatment demonstrated a biphasic activation and increased numbers of CD14+CD16+ monocytes first at the time of infection and later with the emergence of AIDS [282, 306]. Although the SIV-infected macaque model of neuroAIDS best recapitulates HIV neuropathogenesis, there are limitations associated with it. Not all of the SIV infected macaques predictably develop CNS disease – it is suggested that only 25% of these animals develop CNS infection. Secondly, the prolonged progression (1 to 3 years) towards AIDS development limits its usefulness. This has led to other macaque models that take a prolonged time for disease development and progression towards AIDS [267, 307]. The pigtailed macaques co-inoculated intravenously with a combination of two viruses: SIV/17E-Fr (neurovirulent) and SIV/DeltaB670 (immunosuppressive and neurovirulent virus swarm) resulted in depletion of CD4+ T-lymphocytes and efficient virus replication in the CNS macrophages within three months post-inoculation in almost 90% of inoculated macaques [295, 308]. In these models, virus load in the CSF during terminal infection and the CSF:plasma ratio of monocyte chemoattractant protein (MCP)-1 determines the severity of encephalitis.

11.4. SIV and recombinant SHIV challenge strains

It is evident that not all macaque models of SIV infection accurately replicate HIV-1 infection due to several factors, including TRIM5α-mediated restriction. Additionally, the SIV strains also differ from HIV-1 in several ways. For instance, SIVsmm and SIVmac have an accessory gene, vpx (replaces vpu gene), that is not present in HIV-1. HIV-1 and SIVsmm/SIVmac, however, have 53% nucleotide identity. As a result, there has been significant interest in engineering minimally modified HIV-1 strains capable of replicating and recapitulating disease characteristics in macaques.

11.5. SIVs

SIV was first isolated from rhesus macaques with a virulent form of immunodeficiency characterized by opportunistic infections and tumors at the New England Primate Research Center, Southborough, Massachusetts, USA [309]. However, the source of the virus in macaques was later traced to an outbreak of lymphoma in the 1970s at the California National Primate Research Center in Davis, California, USA [310]. Both SIVmac251 and smE660 are independent viral isolates that are commonly used as uncloned viruses. Infectious molecular clones of SIV were derived to obtain genetically defined challenge viruses and proviral DNAs that were suitable for genetic manipulation [311]. SIVmac239 and SIVsmE543-3 are pathogenic molecular clones related in their passage history to SIVmac251 and SIVsmE660, respectively [312, 313]. These viruses’ unique characteristics are that they use CCR5 as a co-receptor, replicate predominantly in memory CD4+ T cells, and express Env glycoproteins that are resistant to neutralizing antibodies. The susceptibility of SIVsmE543-3, and SIVsmE660, to certain rhesus macaque TRIM5 proteins, is an essential factor to be considered in using these challenge viruses. Several investigators are now working to adapt SIVsmE543-3 and SIVsmE660 to restrictive rhesus macaque TRIM5 variants to generate challenge viruses with more uniform resistance in this model [314].

12. Drugs of abuse and SIV

Different drugs of abuse, including opiates, have been found to potentiate HIV replication by diminishing the host-immune responses and accelerating disease progression. For example, SIV-infected Indian Rhesus macaques receiving morphine showed diminished antibody responses, which resulted in severe disease progression, decreased viral clearance rate, increased viral load, and marked depletion of CD4+ T cells [315]. In another study, it was observed that morphine-dependent SIV-infected Indian Rhesus macaques showed enhanced mortality rates, reduced weight gain, higher numbers of circulating CD4+ and CD8+T, higher levels of CSF and plasma viral load, and increased influx of infected monocyte/macrophages into the brains compared to the SIV alone group not dependent on morphine [300]. Indian Rhesus macaques chronically infected with SIVmac239 with/without opioid dependence demonstrated higher viral titer in the presence of morphine compared to the SIV group not administered morphine. An interesting finding from this study was that PBMCs isolated from non-infected animals exposed to morphine resulted in the increased expression of β-chemokine receptor 5 (CCR5). Since both SIV and HIV require CCR5 for cell entry, the unique role of morphine in accelerating viral infection was attributed to morphine-mediated induction of CCR5 expression [316]. Another interesting study demonstrated that morphine-exposed SIVmac239-infected cells (hybrid human B and T cell-line) accumulated in the G1 phase, thus indicating cell cycle arrest. Additionally, morphine was shown to elevate the PKC activity levels and p-ERK1/2, thus implicating that the calcium-PKC-MAPK cascade was involved in morphine-mediated survival of SIV-infected cells [317]. A fascinating study has shown that morphine-dependent (26 weeks), SIV-infected (33 weeks) Indian Rhesus macaques showed a decrease in motor skill tasks in comparison to the morphine or SIV groups alone [318], indicating thereby that morphine could potentiate behavioral impairments in SIV infection. SIV-infected, morphine-dependent macaques also demonstrated significant time-dependent changes in their gut microbial communities (16S) and metabolic profiles. Altered metabolites such as bile acids, sphingolipids, and serotonin could negatively impact the host [319]. It was also shown that there was increased expression of miR-29b in the basal ganglia of morphine-dependent SIV-infected Indian Rhesus macaques compared with the SIV-infected controls and that this involved downregulation of the miR-29b target gene- Platelet-derived growth factor (PDGF)-B, which is a growth factor for neuronal survival [320].

It must be noted that studies on the role of opioids in SIV-mediated immune responses and neuropathogenesis have yielded conflicting results. Some studies have shown that morphine administration prevented developing enzyme-linked immunosorbent spot-forming cell-mediated immune responses in R17/71E-infected rhesus macaques [289]. In contrast, SIV R17/77E infected animals developed cell-mediated immune responses [289]. Additionally, in these animals, morphine treatment did not affect viremia, cerebrospinal fluid viral titers, or survival throughout the study, and the drug was shown to potentiate a more significant build-up of the virus in the brains of infected animals. Furthermore, histopathological changes in the brains showed increased demyelination in morphine-dependent, SIV-infected animals compared to SIV alone group [289]. In another study, morphine dependence exerted a protective effect in a SIV_smm9/rhesus monkey model; however, this study lacked control animals, and virus replication was compared with historical controls [321]. Another study from the same group showed that the mixture of 2 SHIVs (SHIV_KU and SHIV_89.6P) and a neuropathogenic SIV (R71/17E) proved to be an excellent model to develop the rapid and reproducible disease in rhesus macaques [322]. Morphine administration in this model showed no alterations in peak viral loads in rhesus macaques; however, there was higher viral replication at six weeks post-infection in the brain and blood [323]. In light of these findings, controlled studies in non-human primates are warranted to study the effects of drugs of abuse on viral pathogenesis.

Morphine and psychostimulants such as cocaine have also been shown to reduce CD8+ T cells during acute and late stages of SIV infection. However, cocaine administration failed to impact CSF or plasma viral load or CSF CCL2 and IL-6 levels. These findings also suggested that there was no effect of cocaine on SIV encephalitis’s severity compared to vehicle-treated macaques. Additionally, cocaine administration failed to impact neuroinflammation or neurodegeneration, as determined by IFN-β, MxA, CCL2, IL-6, TNFα, IFNγ, and indolamine 2,3-deoxygenase mRNA levels. There was also no change in the APP levels in the presence of cocaine. The executive function of inhibitory control was not impaired in cocaine-treated or control animals following SIV infection. However, animals receiving 3.2 mg/kg/day cocaine performed slower in a bimanual motor test. Chronic administration of cocaine in macaques thus inflicted only minor changes in behavior, encephalitis severity, CNS inflammation/neurodegeneration, and virus replication in SIV-infected pigtailed macaques. These findings thus implicate that cocaine likely only exerts modest effects on the progression of NeuroAIDS in HIV-infected individuals [324].

Interesting studies with another drug of abuse- methamphetamine showed that methamphetamine administration increased viral loads in the brains of SIV- infected macaques but did not alter levels of virus in the plasma. Additionally, methamphetamine treatment also resulted in the activation of natural killer cells. Given the prevalence of methamphetamine use in HIV-infected and HIV at-risk populations, these findings reveal the likely deleterious effects of methamphetamine abuse in such individuals [325]. In another study, microglia were isolated from the brains of macaques infected with SIV, followed by treatment with methamphetamine. Results showed that methamphetamine alone altered the expression of different genes associated with immune pathways, inflammation, and chemotaxis. Most interestingly, it was shown that methamphetamine exposure enhanced the expression of viral entry coreceptors - CXCR4 and CCR5. The increase in CCR5 expression was confirmed in the brain and correlated with increased brain viral load. From this study, it was concluded that methamphetamine increased CCR5-expressing cells, which, in turn, was associated with a concomitant increase in the viral load [326]. Another critical study has shown that cytokines such as IL-15 have been upregulated in astrocytes by methamphetamine, resulting in T cells’ proliferation and an inflammatory phenotype in the brain. Therefore, methamphetamine and IL15 could play critical roles in developing and aggravating CNS immune-mediated inflammatory pathology in HIV-infected drug abusers [327]. Most of the studies have been performed using Indian Rhesus macaques, and the results demonstrate that these drugs of abuse facilitate increased viral replication, which exacerbates viral pathogenesis in the CNS and the periphery.

13. Feline immunodeficiency virus (FIV) model

The feline immunodeficiency virus (FIV) model is an animal model of HIV-1 infection that has been useful in understanding the pathogenesis of AIDS [328]. FIV is a naturally occurring retrovirus that infects domestic and non-domestic feline species and causes an immunodeficiency syndrome, along with clinical symptoms that resemble those observed in humans infected with HIV-1, including symptoms such as periodontitis, neurologic dysfunction, encephalitis, and lymphoma [329–333]. Since the model was developed in 1987, it has proven helpful in several aspects. The primary mode of natural FIV transmission is similar to that of HIV-1 in humans, such as blood-borne infection facilitated by fighting and biting and transmission through transmucosal routes (rectal and vaginal mucosa and perinatal). Furthermore, oral opportunistic infections and changes in the salivary/oral microbiota are also prevalent in FIV infection, and these are associated with the development of oral inflammatory lesions [334–337]. Several early investigations demonstrated that FIV infection results in progressive CD4 T-cell depletion with a primary immune deficiency that causes opportunistic infections, the hallmark of HIV AIDS [328, 329, 331, 338–341]. According to the type and severity of the clinical signs of these infections, FIV disease progression can be divided into four or five stages, described as acute or primary, chronic asymptomatic, persistent generalized lymphadenopathy (PGL), AIDS-related complex (ARC), and feline AIDS (FAIDS), similar to that observed in HIV infection in humans. Intriguingly, FIV infection exhibits a prolonged asymptomatic phase that can last ten years or more, much like HIV-1 infection in humans [331, 342].

FIV is used as a model to study HIV pathogenesis, including immune dysfunction and neurological deficits, and is useful for exploring vaccine development. FIV infection results in acute viremia-associated immune system dysfunction such as CD4 T-cell depletion, peripheral lymphadenopathy, disruption of immune cell function, neutropenia, deterioration of major lymphoid tissues and organs of the hosts, as well as chronic persistent and opportunistic infections due to immunosuppression [330, 338, 343]. Studies on this model have shown that FIV enters the CNS during the acute stage of infection, either via the trafficking of infected monocytes and lymphocytes or by the penetration of free virus across the blood-brain or blood-CSF, barriers similar to HIV-1 [67, 343, 344]. Intriguingly, FIV can infect cells such as microglia and astrocytes in the CNS, and these infected cells serve as a reservoir for latent viral persistence, similar to HIV. The fundamental neurodegenerative and neuropathologic events, including proliferation and activation of glial cells and neuroinflammation, myelin degradation, and neuronal injury/loss observed in HIV, are replicated in the FIV infectious model [344–348]. FIV-infected animals also exhibit the clinical manifestations of the neuropathology of dementia and cognitive-motor processing deficits often observed in HIV infection. This makes FIV a useful animal model to investigate the behavioral pathogenesis seen with retrovirus infection of the CNS [347, 349]. This model has also proven useful in developing and testing neuroprotective therapeutics such as neurotrophin ligands, which prevent the delayed accumulation of intracellular calcium and the decreased cytoskeletal damage of neuronal dendrites [345, 350]. The FIV model is well suited for investigating the effects of drugs of abuse on HIV-associated neuropathogenesis since these models has several commonalities with HIV infection as mentioned above, including 1) the common structural and biochemical properties of FIV and HIV; 2) similar clinical syndromes; 3) a reproducible disease model, with an accurate recapitulation of general and neurological disease progression; 4) FIV infects the CNS and shows neuropathology including neurodegeneration and cognitive and motor dysfunctions [351]. This model was used in previous reports to assess morphine-mediated viral progression [352], behavioral and physical signs of opiate exposure, and evidence of withdrawal symptoms [353]. Morphine-exposed FIV-infected cats failed to exhibit enhanced viral load or severity of FIV-related disease. Morphine demonstrated a protective effect on FIV-associated changes in brain stem auditory evoked potentials. However, morphine-treated cats exhibited clear behavioral and physical signs of opiate exposure and evidence of withdrawal symptoms such as increased cortisol levels [352–354]. In another study, the FIV model was used to explore the potential effects of methamphetamine on lentivirus disease progression, and it was reported that felines are susceptible to low doses of methamphetamine and drug exposure increased FIV replication, as well as the selective and chronic chemical neurotoxicity and disease progression [355]. In another study, FIV-infected cats showed alterations in brain metabolites, such as elevated gamma-aminobutyric acid (GABA) upon methamphetamine exposure [356]. The FIV model is thus suitable for studies of pathogenesis, drug addiction, withdrawal effects, and the development of therapeutics in the context of human HIV infection. However, ethical challenges in terms of animal use and cost have limited this model’s extensive use. Furthermore, this model is not suitable for vaccine development studies and is not closely related to HIV genetic makeup, unlike non-human primate models [357]. The longer incubation periods before the onset of overt disease, less genetic overlap with the HIV genome, and the relatively poor knowledge of the feline immune system are the major factors that limit the applications of the FIV animal model. Importantly, FIV has does not have certain accessory genes that are seen in HIV-1 [329]. Unlike HIV, FIV utilizes CD134 instead of CD4 as the main binding receptor, thereby infecting B cells and CD8+ T cells and CD4+ T cells and macrophages [358, 359]. Since the infection target is different in FIV than HIV/SIV, the reservoir is also different. These factors limit the use of the FIV animal model for cure research.

14. Conclusion and future perspectives

The development of animal models that address HIV and drug abuse comorbidity continues to be an ongoing effort to improve the utility of the existing models. Here we reviewed the existing animal models, discussed their advantages and disadvantages, and learned lessons that could shed light on advancing the field. Chronic low-grade neuroinflammation observed in HIV-positive, drug-abusing humans have been recapitulated in these rodent models. Interestingly, HIV-induced behavioral changes in cognitive performance are also seen in models such as the HIV-1Tg rats, suggesting that viral proteins contribute to cognitive decline and functional deficits, which could be exacerbated by drugs of abuse. Although these rodent models have provided a vast amount of information on mechanisms of HIV-1 and drug abuse interactions, validation in non-human primate models and eventually in human clinical studies is essential. The SIV and SHIV macaque models have improved our understanding of the interaction of HIV-1 and drug abuse. Being primates, they offer the advantage of similarity with humans in immune architecture and response to HIV-1 infection. The heterogeneity and variation in genotypes of macaques, however, poses a challenge, but at the same time also offers an opportunity to examine mechanisms and test therapeutics in a system that most accurately replicates the human disease. Consequently, further development and refinement of these models are required to understand better the effect of drug abuse comorbidity on HIV-1 infection.

Table 1.

List of animal models used in NeuroHIV research.

Model	Mutation/changes	Expression / Target organ	Phenotype	Applications	Advantages	Limitations	Drug abuse
iTat mice [50, 130, 131, 360].	Tat protein expressed in the brain astrocytes under inducible promoter.	Brain–astrocytes.	Increased astrocytosis, loss of neuronal dendrites, and neuroinflammation.	To study the effects of Tat on astrocytes including the astrocyte-mediated Tat neurotoxicity, Tat-impaired neurogenesis, Tat-induced loss of neuronal integrity, and exosome-associated Tat release and uptake.	Depletion of T cells is not necessary to create these models.	Expressed only in brain astrocytes, Expression of tat protein is driven by doxycycline, which has neuroprotective property [131, 360]. Doxycycline inducible promoter is leaky, so small amounts of Tat protein is expressed in early stages of development, because of this leakiness an immune response did not seen after tat induction [158].	Methamphetamine [140], cocaine [141, 361, 362], Morphine [142–144]
EcoHIV mice [61, 151]	The coding region of gp120 in HIV virus was replaced with mouse leukemia virus gp80 to infect mice.	Immune Cells, Spleen, and Brain.	EcoHIV-infected animals showed more macrophages and monocytes.	In the absence of T cells, HIV-infected macrophages are sufficient for induction of cognitive impairment in mice. Good experimental model for NeuroAIDS Therapy.	EcoHIV is a good model to study some of the neuron type-specific neurotoxic functions of HIV. EcoHIV constitutively expressed in mice for up to 16 months.	EcoHIV enters cells through CAT-1 receptor [61].	Methamphetamine [160], Morphine [154].
gp120 transgenic mice [167]	A transgenic gp120 mice model was generated by inserting the portion of the HIV env gene that encodes gp120 into the mouse genome under the control of a glial fibrillary acidic protein (GFAP) promoter, resulting in gp120 protein expression by brain astrocytes.	Brain-astrocytes.	Abnormal dendrite morphology, decreased synaptic and dendritic density, Increased astrocytosis.	Good animal model for neuroAIDS research	Non-infectious model, Neurodegene ration and glial cell activation mimics that of HIV encephalitis. Protein express only in brain astrocytes.	Viral protein is present throughout the lifespan might promote survival of abnormally robust neuronal populations [158].	Methamphetamine [174–176], Morphine: [138, 172].
SCID [191, 363-366]	Mutation in the gene for protein kinase, DNA activated, catalytic polypeptide (PRKDC).	Thymus, spleen, lymph node.	Absence of functional B cells and T cells.	To study the human immune system and human disease in a small animal model.	Useful tool for Immunology, Inflammation and Autoimmunity Research, Xenograft/transplant host AIDS research tool.	In certain circumstances like unsterile conditions, can lead to activation of NK cells, leakiness of SCID mice.[195, 196]	Morphine [191], Methamphetamine [192, 193], cocaine: [194, 367].
Tg 26 Mice [103, 108]	Loss of neuronal dendrites, loss of synapses. Kidney disease, HIV-associated nephropathy. Tg26 mice exhibit proteinuria.	Ubiquitous	Loss of neuronal dendrites, loss of synapses. Kidney disease, HIV- associated nephropathy. Tg26 mice exhibit proteinuria.	Animal model for human HIV-associated nephropathy (HIVAN).	HIV-1-Tg26 mouse model has been used extensively to study the pathogenesis of HIVAN because these mice develop renal disease mimicking human HIV associated Nephropathy.	Approximately 15% of Tg 26 mice spontaneously develop leukemia/lymphoma [368].	Cocaine [116], Morphine [117], Methamphetamine [173, 369].
HIV transgenic rats [56, 220, 370]	Plasmid containing non-infectious HIV provirus and host cell flanking regions were microinjected into fertilized rat eggs.	lymph nodes, thymus, liver, kidney, and spleen.	Cataracts, Kidney and heart damage and loss of lymphocytes.	Good model for neuroHIV study.	Transgenic rat is a noninfectious small-animal model of HIV-1 pathogenesis because it efficiently express viral gene in lymph nodes, spleen, thymus, and blood. Useful to study the mechanisms underlyin HIV - associated disease progression caused by HIV proteins.	This model is not useful to study the initial infection stage of HIV and there is no viral replication in the HIV-1Tg rat [228].	Methamphetami ne[224, 371, 372], Morphine, [225, 226], Cocaine: [227].
SIV and SHIV monkey models [235–238, 309]	Intravenous inoculation of SIV virus.	Brain, Heart, and Kidney.	Reduced number of CD4+ T cells.	Good model system for HIV pathogenesis and vaccine research.	SIV is genetically, antigenically and morphologically close to HIV. SIV infected monkeys are immunosuppressed and display neuropathological and behavioral features are identical to HIV-infected humans.	higher maintenance costs, Not susceptible to HIV-1 [240].	Morphine [239, 315, 316], Cocaine [324], Methamphetami ne[325–327].
FIV model [328]	FIV is a natural lentiviral. The integrated provirus possesses gag, pol, and env genes, common elements of all retroviruses.	Liver, Kidney, Brain and Lung.	Progressive CD4+ T cell depletion, a deficit in the humoral immune response, and the dysregulation of cytokine expression.	Valuable animal model for the development of antiretroviral drugs.	Not large in size and not difficult to handle, The models can be infected with FIV, FeLV and develop immuno deficiency-like syndrome.	Cat’s genetic makeup is far related to human FIV, FeLV are not closely related to HIV [330, 357].	Methamphetami ne[354, 355], Morphine[352, 353],
scid-hu-Thy/Liv mice. [373, 374]	SCID mice are implanted with human liver and thymus.	Thymus and Liver	Decrease in the CD4/CD8 ratio	This model is useful to study HIV pathogenesis and to test the efficacy of antiretroviral drugs for HIV replication.	These mice have human liver and thymus cells, which are the target organ for antiretroviral drug uptake and action.	Substantial degree of ‘leakiness’ in certain mouse strains (leading to the development of mouse T cells and B cells in older animals. This model cannot be used to study mucosal transmission of HIV-1 [240].	N/A
Humanized microglial mice. [216, 375, 376]	Mice express a humanized form of the microglial growth factor CSF1, which is essential for microglial maintenance, and lack of two immune factors required for rejection of foreign cells.	Brain	N/A	This model will be useful to investigate how human and mouse microglia function differently in normal and HIV conditions. Useful to study the role of human microglia in brain development and degeneration.	Useful to study the pathophysiology of the human cells within an intact brain.	Many human cytokines and other factors are species specific [216, 376] .	N/A

Highlights:

• HIV-1 infection and drug abuse are intertwined epidemics.

• Animal models are necessary to decipher the impact of drugs on HAND progression.

• Non-human primates are gold-standard model in this study but very expensive.

• Several strategies have also been used to develop several rodent models of HAND.

• These rodent models serve the purpose of modeling some unique aspects of HAND.