Characterization of Macrophage-Tropic HIV-1 Infection of Central Nervous System Cells and the Influence of Inflammation

Abstract

HIV-1 infection within the central nervous system (CNS) includes evolution of the virus, damaging inflammatory cascades, and the involvement of multiple cell types; however, our understanding of how Env tropism and inflammation can influence CNS infectivity is incomplete. In this study, we utilize macrophage-tropic and T cell-tropic HIV-1 Env proteins to establish accurate infection profiles for multiple CNS cells under basal and interferon alpha (IFN-α) or lipopolysaccharide (LPS)-induced inflammatory states. We found that macrophage-tropic viruses confer entry advantages in primary myeloid cells, including monocyte-derived macrophage, microglia, and induced pluripotent stem cell (iPSC)-derived microglia. However, neither macrophage-tropic or T cell-tropic HIV-1 Env proteins could mediate infection of astrocytes or neurons, and infection was not potentiated by induction of an inflammatory state in these cells. Additionally, we found that IFN-α and LPS restricted replication in myeloid cells, and IFN-α treatment prior to infection with vesicular stomatitis virus G protein (VSV G) Envs resulted in a conserved antiviral response across all CNS cell types. Further, using RNA sequencing (RNA-seq), we found that only myeloid cells express HIV-1 entry receptor/coreceptor transcripts at a significant level and that these transcripts in select cell types responded only modestly to inflammatory signals. We profiled the transcriptional response of multiple CNS cells to inflammation and found 57 IFN-induced genes that were differentially expressed across all cell types. Taken together, these data focus attention on the cells in the CNS that are truly permissive to HIV-1, further highlight the role of HIV-1 Env evolution in mediating infection in the CNS, and point to limitations in using model cell types versus primary cells to explore features of virus-host interaction. IMPORTANCE The major feature of HIV-1 pathogenesis is the induction of an immunodeficient state in the face of an enhanced state of inflammation. However, for many of those infected, there can be an impact on the central nervous system (CNS) resulting in a wide range of neurocognitive defects. Here, we use a highly sensitive and quantitative assay for viral infectivity to explore primary and model cell types of the brain for their susceptibility to infection using viral entry proteins derived from the CNS. In addition, we examine the ability of an inflammatory state to alter infectivity of these cells. We find that myeloid cells are the only cell types in the CNS that can be infected and that induction of an inflammatory state negatively impacts viral infection across all cell types.

Keywords: HIV-1; HIV-1 CNS inflammation; HIV-1 Env evolution; HIV-1 astrocyte infection; HIV-1 microglia infection; HIV-associated neurocognitive disorders; macrophage-tropism; neuroHIV.

FIG 1

Pseudotyped viruses with Env proteins derived from uncultured virus display CCR5 or CXCR4 coreceptor usage. (A) Affinofile cells induced to express high (H, striped) or low (L, striped) levels of CD4 and high levels of CCR5 were infected with HIV-1 luciferase reporter viruses pseudotyped with patient-derived macrophage-tropic (blue), T cell-tropic (red), or X4 T cell-tropic (orange) envelope proteins. Negative-control viruses (cyan) lacked an envelope protein, whereas the positive-control viruses (pink) were pseudotyped with vesicular stomatitis virus G protein, which is capable of infecting a wide range of mammalian cells. Infection was measured in relative light units (RLUs) via luminometer. (B) Affinofile cells induced to express high levels of CD4 and low levels of CCR5 were treated with 20 nm, 200 nm, or 2000 nm of Maraviroc (M20, M200, M2000) or AMD3100 (A20, A200, A2000) and then infected with HIV-1 luciferase reporter viruses pseudotyped with patient-derived macrophage-tropic (blue), T cell-tropic (red), or X4 T cell-tropic (orange) envelope proteins. Negative-control viruses (cyan) lacked an envelope protein, whereas the positive-control viruses (pink) were pseudotyped with vesicular stomatitis virus G protein, which is capable of infecting a wide range of mammalian cells. Infection was measured in RLUs via luminometer.

FIG 2

Macrophage-tropic HIV-1 Env proteins confer an entry advantage over R5 T-tropic Env proteins, and LPS and IFN restrict infection in monocyte-derived macrophages (MDMs). MDMs were treated with 10 ng/mL (L10) or 100 ng/mL (L100) lipopolysaccharide (LPS) or interferon alpha (IFN-α) for 24 h and then infected with HIV-1 luciferase reporter viruses pseudotyped with patient-derived R5 macrophage-tropic (blue), R5 T cell-tropic (red), or X4 T cell-tropic (orange) envelope proteins. Untreated control groups are labeled “Con” within each Env group. Negative-control viruses (cyan) lacked an envelope protein, whereas the positive-control viruses (pink) were pseudotyped with vesicular stomatitis virus G protein, which is capable of infecting a wide range of mammalian cells. The average magnitude difference between the positive and negative control was 6,066×. Infection was measured in relative light units (RLUs) via luminometer. N = 6 technical replicates.

FIG 3

LPS and IFN restrict HIV-1 infection in primary and iPSC-derived microglia, but cell models are not equally permissive. (A and B) Primary human microglia (A) and human iPSC-derived microglia (B) were treated with 10 ng/mL (L10) and 100 ng/mL lipopolysaccharide (LPS) or interferon alpha (IFN-α) for 24 h and then infected with HIV-1 luciferase reporter viruses pseudotyped with patient-derived R5 macrophage-tropic (blue), R5 T cell-tropic (red), or X4 T cell-tropic envelope proteins. Untreated control groups are labeled “Con” within each Env group. Negative-control viruses (cyan) lacked an envelope protein, whereas the positive-control viruses (pink) were pseudotyped with vesicular stomatitis virus G protein, which is capable of infecting a wide range of mammalian cells. The average magnitude difference between the positive and negative control was 106× for primary microglia and 113,127× for iPSC-derived microglia. Infection was measured in relative light units (RLUs) via luminometer. N = 3 technical replicates.

FIG 4

Astrocytes are not infected by macrophage-tropic or R5 or X4 T cell-tropic HIV-1, and inflammation does not potentiate infection. (A to C) Primary astrocytes (A), transformed human astrocytes (B), and human iPSC-derived astrocytes (C) were treated with 10 ng/mL and 100 ng/mL lipopolysaccharide (LPS) or interferon alpha (IFN-α) for 24 h and then infected with HIV-1 luciferase reporter viruses pseudotyped with patient-derived R5 macrophage-tropic (blue), R5 T cell-tropic (red), or X4 T cell-tropic envelope proteins. Untreated control groups are labeled “Con” within each Env group. Negative-control viruses (cyan) lacked an envelope protein, whereas the positive-control viruses (pink) were pseudotyped with vesicular stomatitis virus G protein, which is capable of infecting a wide range of mammalian cells. The average magnitude difference between the positive and negative control was 790× in primary astrocytes, 164,689× in transformed astrocytes, and 351,361× for iPSC-derived astrocytes. Infection was measured in relative light units (RLUs) via luminometer. N = 3 technical replicates.

FIG 5

Neurons are not infected by macrophage-tropic or T cell-tropic HIV-1, and inflammation does not potentiate infection. (A and B) Primary human neurons (A) and transformed human neurons (B) were treated with 10 ng/mL and 100 ng/mL lipopolysaccharide (LPS) or interferon alpha (IFN-α) for 24 h and then infected with HIV-1 luciferase reporter viruses pseudotyped with patient-derived R5 macrophage-tropic (blue), R5 T cell-tropic (red), or X4 T cell-tropic envelope proteins. Untreated control groups are labeled “Con” within each Env group. Negative-control viruses (cyan) lacked an envelope protein, whereas the positive-control viruses (pink) were pseudotyped with vesicular stomatitis virus G protein, which is capable of infecting a wide range of mammalian cells. The average magnitude difference between the positive and negative control was 1514× in primary neurons and 642× in transformed neurons. Infection was measured in relative light units (RLUs) via luminometer. N = 3 technical replicates.

FIG 6

Inflammation induces modest changes in HIV-1 entry receptor transcript expression. (A and B) Following treatment, mRNA sequencing was performed in parallel, noninfected samples to determine basal expression of cell-type-specific markers (A) and HIV entry receptors (B), as well as changes in HIV entry receptor transcripts in response to a 24-h IFN-α or LPS treatment. Counts per million (CPM) were generated in Partek Flow for each gene by normalizing for total reads per samples. (A) The mean CPM (n = 3) and standard deviation for each gene are listed for tASTRO (light blue), pASTRO (light purple), tNEU (pink), pNEU (green), MDM (teal), iMGL (dark purple), and pMGL (peach). A) CPM values were generated for P2RY12 (microglia-specific), CD11b (myeloid-specific), NeuN (neuron-specific), and GFAP (astrocyte-specific) to validate the identify of transformed and primary cell types. (B) CPM values were generated for HIV entry receptor CD4 and coreceptors CCR5 and CXCR4. Cell type colors are identical to panel A. (C) Differential expression analysis was performed using DESeq2, and log₂ fold changes (padj < 0.05) are listed for CD4 (left), CCR5 (middle), and CXCR4 (right) transcripts for tNEU (orange), MDM (purple), pMGL (blue), and iMGL (pink) following LPS or IFN treatment at 10 ng/mL (LPS1/IFN1) or 100 ng/mL (LPS2/IFN2). Transformed astrocytes, primary astrocytes, and primary neurons did not demonstrate changes in HIV-1 entry receptor expression and thus do not appear on this graph.

FIG 7

Myeloid-derived cells express similar transcriptomes, but astrocyte and neurons differ significantly by cell type and model. A principal-component analysis (PCA) was generated in Partek Flow using all normalized reads from treated and nontreated pMGL (blue), iMGL (red), MDM (yellow), tASTRO (purple), pASTRO (green), tNEU (pink), and pNEU (teal). Individual replicates for treated (IFN1, squares; IFN2, crosses; LPS1, Xs; LPS2, stars) and nontreated (circles) are displayed for each cell type. Total variance comprised by vector is listed on the correlating axis. (A and B) PC1 versus PC2 is best visualized in a head-on view (A) whereas rotation of the three-vector plot reveals the influence of PC3 (B).

FIG 8

CNS cell types demonstrate nonuniform responses to IFN- and LPS-mediated inflammation. (A) Hierarchical clustering was performed using on the log₂ fold change of genes across all treatments. Fold changes compared to nontreated samples (upregulation, yellow; downregulation, purple) are listed for each gene across all cell types. Treatment groups include IFN1 (10 ng/mL; green), IFN2 (100 ng/mL; purple), LPS1 (10 ng/mL; orange), and LPS2 (100 ng/mL; yellow). Only genes with a padj of <0.05 and fold change of ±4 in at least one analysis were included. Cluster distances are based on correlation distance with complete linkage. (B) Total differentially expressed genes (padj < 0.05, fold change of ±2) in response to IFN2 and LPS2 are plotted for each cell type. Differentially expressed genes shared between all cell types are listed below the histograms (LPS, 0; IFN, 57). (C) Gene pathway enrichment analysis was performed using Enrichr for the 57 genes commonly differentially expressed across all cell types in response to 100 ng/mL IFN. Log₁₀ P values representing the strength of correlation are listed for each pathway.

FIG 9

MDMs and pMGL share 1,017 differentially expressed genes across IFN-α and LPS treatments. A Venn diagram was generated to compare differentially expressed genes across all MDM and pMGL treatment groups. The number in each section represents the number of differentially expressed genes shared by the overlapping cell types/conditions, while the percentage represents the group as a fraction of total differentially expressed genes across the analysis.