Human Microglia Models for NeuroHIV

1. Introduction

As the primary immune cells in the brain, microglia play a crucial role in the brain's immunity against viral infections, including HIV [1]. However, because microglia express the HIV entry receptors (CD4, CCR5 and CXCR4), they also serve as a major target for both productive and latent HIV infection in the brain [2,3,4]. Importantly, HIV-infected microglia can produce cytokines, reactive oxygen species, and the viral products which create an inflammatory environment in the brain, leading to neurodegeneration and HIV-associated neurocognitive disorder (HAND) [5]. Therefore, it is essential to understand the role of microglia in the immunopathogenesis of HIV infection in the brain. However, studies with primary human microglia have been hampered by the difficulties of obtaining surgical human brain specimens, isolating high quality/quantity of microglia and the cells’ short lifespan in culture. To overcome these issues, researchers have developed various human microglial culture models for NeuroHIV studies (Table 1). This review focuses on these models that have been commonly used by different laboratories, including primary human microglia, microglial cell lines, peripheral blood monocyte-derived microglia (MMG) or MMG-like cells, HIV-latently infected microglial cell line, human induced pluripotent stem cell (iPSC)-derived microglia (iMg), and microglia containing cerebral organoids (MCOs) from iPSC.

2. Primary Human Microglia

Primary human microglia can be isolated from the brain tissue of a deceased person and cultured for several weeks under microglia-specific enrichment conditions [6,7]. Other sources of primary human microglia are fetal brain tissues following abortions and patients undergoing surgery for brain tumors or epilepsy [8,9,10]. The technique protocols for such isolation have been well-established [7,11,12,13]. Like primary human macrophages derived from peripheral blood monocytes, in vitro cultured microglia are highly susceptible to HIV infection [14], and infected cells produce inflammatory cytokines [2,15]. However, obtaining fresh human brain tissue to isolate large quantities of viable microglia is extremely challenging. Additionally, the heterogeneity of samples, obtained from different brain regions of different donors (differing in age, sex, health condition, etc) can result in many confounding variables for experiments with primary human microglia. Therefore, it is crucial to develop alternative in vitro human microglia models for HIV studies.

3. Microglial Cell Lines (HMC3 and HMO6)

HMC3 resembles the adhering characteristics of primary human microglia but has low phagocytic activity [16]. The cells express some of the microglial markers and antigenic molecules such as IBA-1, CXCR1, CXCR3, NSE, and MHC-II/IFN-ɣ. However, they do not express CD14, CD68/Ki-M6, CD11c and P2RY12 [17,18,19]. Importantly, HMC3 does not express the major HIV entry receptor CD4 and is not permissive for productive HIV infection/replication. Studies of HIV infection with HMC3 are only limited to the experiments with the pseudo-typed viruses [17,18]. Moreover, it has been reported that these cells could be of Rat origin [20,21]. Nagai et al., developed another immortalized microglial cell line, named HMO6. Although HMO6 expresses the microglial markers (CD11b, CD68, CD86, HLA-ABC, HLA-DR, and RCA-1 Lectin), these cells have a less diverse profile of secreted soluble inflammatory mediators than primary human microglia [22]. Like HMC3 cells, HMO6 lacks the expression of microglial markers (CD14 and CD11c) and CD4 receptor [23], and thus their use for HIV studies is limited [6,22,23,24].

4. Microglial Cell Line for Latent HIV Infection (C20 and H69.5)

Garcia-Mesa et al. [21] developed an immortalized microglial cell line (C20) for HIV studies. C20 was generated from primary glia isolated from human adult brain tissues and frozen glial cells. The cells have microglia-like morphology and express the key microglial markers (CD11b, TGFβR, and P2RY12). Importantly, their RNA expression profiles have similar characteristics of primary human microglial cells. C20 expresses CCR5 but has extremely low level of CD4 receptor expression which decreases with increasing cell passages. Therefore, the cells are not suitable for productive HIV infection [21]. However, David Alvarez-Carbonell et al. [25] used C20 as the parental cell line to establish HIV latently infected cell line (HC69.5). Briefly, HC69.5 was developed by immortalizing C20 cell line with simian virus 40 large T antigen/human telomerase reverse transcriptase [25]. These immortalized cells were then transfected with a vesicular stomatitis virus G envelope pseudo-typed lentiviral vector with a green fluorescent protein as a reporter [25]. HC69.5 expresses the specific microglial markers (P2RY12 and CD11b) and the macrophage lineage marker (CD14). Importantly, HC69.5 cells express the HIV RNAs, and the viral proteins (Tat, Rev, Env, Vpu and Nef), but lack HIV gag [21,26]. HC69.5 has been widely used as a latently infected HIV microglia model which could be significantly activated by TNF-α [27,28]. Studies show that these cells express low levels of HIV (1-6%), but TNF -α induction for 16h could significantly induce HIV replication in 90% of the cell population [25]. Additionally, autocrine expression of TNF-α can spontaneously reactivate HIV in HC69.5 cells, which can be blocked by the treatment of glucocorticoid receptor agonist, dexamethasone [27,28].

5. Human Peripheral Blood Monocyte-Derived Microglia (MMG)

Rawat et al. developed a protocol for generating peripheral blood monocyte-derived microglia (MMG) [29], which was based on the early work MMG-like cells by Leone et al. [30], they demonstrated that human peripheral blood monocytes cultured in serum-free medium with M-CSF, GM-CSF, NGF, and CCL2 could differentiate into MMG. These cells are different in morphology, phenotype, and function from freshly isolated monocytes. They resemble primary human microglia and express microglia markers, including CD11b, CD11c, CD14, CD45, CD195, CD80, P2RY12, IBA-1, CD14, and CD45. MMG also express low levels of HLA-DR, and CD86. Importantly, MMG expresses the key HIV entry receptors (CD4, CCR5, and CXCR4) [18] and could be productively infected with HIV, and support long-term infection while continuously releasing virions into the culture media [18,29].

The reported protocols for generating MMG are relatively simple, feasible and reliable. Ohgidani et al. showed that the addition of microglial-growth factors (GM-CSF and IL-34) to the human monocyte cultures for 2 weeks, can convert monocytes to microglia like cells, which express CD11b high/CD45low and CX3CR1high/CCR2low [31]. These cells released pro-inflammatory/ anti-inflammatory cytokines and can perform phagocytosis [31]. Subsequent studies on MMG showed that additional factors, such as M-CSF, nerve growth factor (NGF)-β, and CC chemokine ligand 2 (CCL2), were suitable for HIV infection [29,32]. Sheridan et al. adapted this protocol for differentiating microglia from umbilical cord blood-derived mononuclear cells [33]. While MMG have been widely used as a microglial model for HIV studies, there is a concern about whether the added growth factors and the cytokines could interfere with the study outcomes, because some of these factors are known to be implicated in HIV infection.

6. Human Induced Pluripotent Stem Cells (iPSC)-Derived Microglia (iMg)

The discovery of induced pluripotent stem cells (iPSCs) has provided a platform for the generation of a wide variety of brain cell types, including human microglial cells (iMg) [34,35,36,37]. iMg are like primary human microglia in morphology, gene expression, and cytokine release profile. They are distinct from other tissue macrophages as they display a profile of neuronal-co-culture-specific expression and inflammatory response. The iMg model has been utilized to study neurological diseases [38], such as Alzheimer’s disease [39], Parkinson’s disease [40], amyotrophic lateral sclerosis, and frontotemporal dementia [41]. Additionally, iMg can be infected by Zika and dengue virus [42]. Importantly, iMg expresses the microglial markers (P2RY12 and TMEM119) and the HIV entry receptors (CD4, CCR5 and CXCR4) [18]. Several groups have reported that iMg could be productively infected with HIV, particularly with CCR5 tropic strains [6,18,43]. We recently reported that iMg possess immunologically functional toll like receptor 3 (TLR3) which could be activated by Ploy (I:C) and produce the antiviral cellular factors against HIV [44]. Although iMg has been successfully used in HIV infection studies, there is little information on establishing persistent/latent HIV infection in these cells, which is likely due to their short in vitro lifespan. The protocols for generating human iMg have been well established [45]. Mcquade et al. published a simplified protocol for establishing iMg [46]. Additionally, several companies have now provided detailed protocols, the culture media and iMg cells originated from different donors (Table 2).

7. Microglia-Containing Cerebral Organoids (MCOs) Derived from Human iPSCs

In the in vivo microenvironment, microglial functions significantly depend on their direct and/or indirect contact with other brain cell types such as neurons and astrocytes. Therefore, it is clinically important to develop a microglia-containing cellular model with other key brain cells. In 2013, Lancaster et al. [47] reported the development of cerebral organoids (COs) from the iPSC. Since then, the field of iPSC-derived COs has been significantly advanced [48,49,50,51,52]. The major advantage of COs is that the cultured cells can self-organize into 3D structures and differentiate into the key major brain cell types, which recapitulate the layered structure, cellular diversity, and synaptic connectivity of the human brain [53,54]. Recently, human iPSC-derived COs have been increasingly used as a brain model for studying various neurological disorders and neurotropic virus infections. Importantly, it has been documented that iPSC-derived COs can be cultured for many months, and the longest duration of maintaining COs in culture was 800 days [55]. This feature of COs allows long-term studies on neurodevelopment or disease progression, which is particularly important for studying latent HIV infection.

Although iPSC-derived COs recapitulates some key features of human brain development, many of the currently used COs are derived from neuroectodermal progenitors and only contain neurons and astrocytes. They do not have microglia which arise from mesodermal progenitors. Therefore, the absence of microglia in COs substantially limits their value and applicability for brain research, particularly HIV studies. As resident immune cells in the brain, microglia are crucial not only for the brain immunity but also for neurogenesis and neuroinflammation. More importantly, microglia are the primary target and reservoir for HIV infection. Park et al. developed microglia-sufficient brain organoids by co-culturing COs with primitive-like macrophages generated from the human iPSC (56). They demonstrated that iPSC-derived microglia promote organoid maturation via cholesterol transfer [56]. We and others have developed the protocol to generate microglia containing cerebral organoids (MCOs) [50,57,58,59], demonstrating that MCOs express microglial markers (P2RY12 and TMEM112) and the major HIV entry receptors (CD4 and CCR5).

HIV infection of the MCOs model was first reported by Dos reis et al [60]. They incorporated HIV-infected primary human microglia or the microglial cell-line (HMC3) into COs. They demonstrated that this model supported low levels of HIV replication and that the HIV-infected microglia can produce inflammatory factors in COs (60). Another group, Gumbs et al. [6] demonstrated that both MCOs and isolated organoid-derived microglia could be productively infected with replication-competent HIV-Bal reporter viruses. They found that the susceptibility of organoids to HIV infection was associated with the expression of the microglia marker (AIF1) and the HIV entry receptors (CD4 and CCR5) regardless of organoid maturation. Other groups also reported that productive HIV infection was only observed in microglial cells which was dependent on the co-expression of microglia-specific markers and the CD4/CCR5 receptors [18,61,62]. More recently, Donadoni et al. showed that HIV replication in MCOs could be inhibited by cART [61]. We have also observed that the MCOs from some human iPSC lines could be acutely infected by live HIV Bal strain (unpublished data). In addition, in agreement with the study by Gumbs [6], we found the significant variability between organoids from the same batch and across iPSC lines in terms of susceptibility to HIV infection. Furthermore, there has been no data showing that MCOs support persistent/latent HIV infection (Table 3). These issues highlight an important limitation of the brain organoid model for HIV infection study.

8. Discussion: Pros and Cons of the Human Microglia Models

In Table 4, we summarize the advantages (pros) and disadvantages (cons) of the human microglia models focusing their feasibility and applicability to HIV acute and persistent infection. Generally, while the immortalized cell lines are readily available and resemble some aspects of the human primary microglia, their usage particularly for HIV infection is limited because they do not express the primary HIV entry receptor, CD4 (Table 1 and Table 4). Among the immortalized cell lines, only HC69.5 has been used as a human microglia model for studying HIV latency [26,27]. In addition to the microglial cell lines, human peripheral blood monocyte-derived microglia (MMG) have been used as a microglia model for HIV infection. However, there is concern about whether the addition of the microglial growth factors and cytokines to the cultures can introduce variables that affect the study outcomes. Some of these factors are known to be implicated in regulating HIV infection/replication and the innate immune function of these cells. Recently, iPSC-derived microglia (iMg) model has gained great attention from investigators because like primary human microglia iMg can be productively infected with HIV [6,18,44]. However, both MMG and iMg are fully differentiated cells with limited lifespan, which is an obstacle for establishing persistent and latent HIV infection. When compared with 2D cultures of the microglia models, 3D brains organoid model, particularly MCOs, is a more advanced in vitro model for studying HIV brain infection and infection-mediated neuropathogenesis. Various groups (Table 3) have shown that MCOs could be productively infected by HIV and produced inflammatory cytokines. However, while MCOs have a significantly longer in vitro lifespan than iMg, there is little information about whether MCOs support persistent/latent HIV infection (Table 4). In addition, both COs and MCOs do not contain the blood-brain barrier (BBB) and the perivascular macrophages, another major target of HIV in the brain. Moreover, the shape, size, and maturing time point of MCOs vary among iPSC line donors and culture batches. During the long-term cultures, the core of COs or MCOs cannot obtain sufficient nutrients/oxygen, resulting in cell death within organoids. Therefore, overcoming these limitations [52] is essential for further improving the brain organoid models for NeuroHIV studies.