HIV-1 infection facilitates Alzheimer’s disease pathology in humanized APP knock-in immunodeficient mice

Study approvals

All animal procedures in these studies followed the Institutional Animal Care and Use Committee (IACUC) guidelines approved protocols at the University of Nebraska Medical Center (UNMC).

Development and generation of NAIL mice

NAIL mice were generated on a NOD. Cg-Prkdc ^scid Il2r ^gtm1Sug /JicTac (NOG) (Taconic, Germantown, NY) background carrying a knocked-in APP^KM670,671NL and transgene for IL34 (NOG/APP^KM670,671NL/[IL34]) (NAIL mice) [31]. Briefly, the human gene containing a familial AD mutation was knocked in by substituting five nucleotides at exon 16 of mouse APP (Figure 1A). The humanized genomic sequences of familial AD mutations were introduced into fertilized eggs of pregnant NOG mice through microinjection. Appropriate guide RNAs were used to remove mouse exons, and the mouse donor DNA was replaced with human mutations through DNA repair. The eggs were then transplanted into pseudo-pregnant recipients, and the offspring were genotyped through gene-specific RT-PCR and Sanger sequencing. Sanger sequencing was used to confirm the knock-in of humanized APP sequences. Ear tissue was collected at weaning and digested using an alkaline lysis buffer 0.4 M NaOH and 0.2 mM ethylene diamine tetraacetic acid (EDTA). Samples were heated at 95 °C for 1 h until fully digested. Samples were then neutralized with a neutralization buffer (40 mM Tris-HCl) and centrifuged at 10,000 g for 5 min. Supernatants were used for DNA amplification. Polymerase chain reaction (PCR) was performed to amplify the gene of interest (APP) using GoTaq Hot Start Green Master Mix (cat. M5123, Promega, Madison, WI) and forward and reverse primers specific to each gene (APP for: GGCGGTCACACTAACGGATG, APP rev: GACACAGGACAAGCCACGAG). PCR products were confirmed by gel electrophoresis on a 1 % agarose gel, then purified using the ZR DNA Sequencing Clean-up Kit (cat. D4051, Zymo Research, Irvine, CA). Following purification, samples were mixed with the forward primer for the gene of interest and submitted to Azenta Life Sciences (Burlington, MA) for sequencing. SnapGene Viewer was used to visualize nucleotide sequences and identify mutations to determine whether mice were heterozygous or homozygous for humanized APP [31].

Figure 1:

Development of AD humanized mouse knock in of a APP (Swe) mutation. (A) Nucleotide and amino acid substitutions in exon 16 provided to introduce the APP Swedish mutation (APP^KM670,671NL) (red highlighted) and humanize the mouse APP exon 16 (green highlighted) as shown by Sanger sequencing. This confirmed replacement of pathological (red box) and humanized (green box) nucleotides in mouse APP gene. (B) Developmental scheme for generating NAIL mice showing (NOD Cg-Prkdc ^scid Il2rg ^tm1Sug /JicTac) (NOG) with knocked in APP^KM670,671NL (NA) mice. Homozygous NA mice were crossed with NOD Cg-Prkdc ^scid Il2rg ^tm1Sug Tg(CMV-IL34)1/Jic that are competent for the development of human-like microglia (Hu Microglia) to generated the NAIL mice. First-generation NAIL mice carry heterozygous human APP and human IL34 genes. Homozygous NA mice were backcrossed with heterozygous NAIL mice to generate second-generation NAIL mice that carry homozygous human APP and heterozygous human IL34 genes with competency for permissive human-like microglia development. (C) qPCR analyses confirmed the presence of the human IL34 gene in NAIL mice. (D) Animal study design showing the timeline of experimental interventions. Neonatal NAIL mice were engrafted with HSCs at 2–3 days after birth and were monitored for human CD45+ immune cell reconstitution for 3 months by peripheral blood flow cytometry. At approximately four months of age, human immune reconstituted NAIL mice were infected with 10⁴ TCID₅₀ of HIV-1_ADA and monitored for infection at 10- and 25 days post-infection. Mice were sacrificed at 25 days following infection. (E) Representative flow cytometric gating and frequencies of human immune cells in peripheral blood of 3-month-old NAIL mice. The presence of human CD45+ immune cells, human T cells (CD3+, CD4+, and CD8+), human B cells (CD19+), and human monocyte/macrophages (CD14+) in NAIL mice confirmed human immune reconstitution of NAIL mice before HIV-1 infection.

APP-KI mice were crossed with another immunodeficient NOG background mice containing human IL34 transgene (NOD Cg-Prkdc^scidIl2rg^tm1SugTg(CMV-IL34)1/Jic) (referred to as hIL34/NOG) to generate NAIL mice (Figure 1B). Since mice are on an immunocompromised NOG background, they allow the reconstitution of human adaptive immune cells. Additionally, human IL34 transgene enables the development of human microglia in the brain after the engraftment of hematopoietic stem cells. The presence of transgenic human IL34 by NAIL mice was compared to background NOG mice using a human IL34 gene-specific primer (Hs01050926_m1). One-to two-day-old neonates were toe-clipped before human hematopoietic stem cell reconstitution to perform qPCR to confirm the presence of the hIL-34 gene in NAIL mice. We used the following real-time PCR thermal cycle setting: 50 °C for 2 min, 95 °C for 15 s, 40 cycles of denaturation at 95 °C for 15 s, and annealing at 60 °C for 1 min. We used GAPDH as the housekeeping gene. A cycle-threshold (CT) value of less than 30 (CT<30) was considered positive for the presence of the human IL34 gene in NAIL mice [32].

Reconstitution of the human immune system in NAIL mice

NAIL mice, 2–3 day-old neonates, were irradiated with 1 Gy (RS 2000 X-ray Irradiator, Rad Source Technologies, Inc., Suwanee, GA, USA). Four hours post-irradiation, neonates received an intrahepatic injection of 10⁵ CD34+ hematopoietic stem cells (HSCs) derived from human cord blood (Lonza Biosciences, Rockland, ME). The purity of the CD34+ HSC (>90 % purity) was confirmed by the supplier (Lonza Biosciences) using flow cytometry. After 3 months of CD34+ cell engraftment, mice were bled from the submandibular vein for flow cytometry to confirm the reconstitution of human peripheral immune cells [38]. Six reconstituted animals were used in the study.

Flow cytometry of human immune cell populations from NAIL mice

Blood samples were collected in EDTA-containing tubes (BD Microtainer, Franklin Lakes, NJ, USA) and centrifuged at 350 g for 8 min. Blood cells were resuspended in a fluorescence-activated cell sorting (FACS) buffer 2 % fetal bovine serum (FBS) and 0.1 % sodium azide in phosphate-buffered saline (PBS), then incubated for 30 min at 4 °C with a cocktail of antibodies against human immune cell markers that included fluorescein isothiocyanate (FITC)-anti-CD45, Alexa Fluor 700 (AF700)-anti-CD3, phycoerythrin-cyanin 5 (PE-Cy5)-anti-CD19, allophycocyanin (APC)-anti-CD4, Brilliant Violet 421 (BV421)-anti-CD8, and phycoerythrin (PE)-anti-CD14. All antibodies and isotype controls were obtained from BD Biosciences (San Jose, CA). Red blood cells were lysed by FACS lysing solution (BD Biosciences). Stained cells were washed with FACS buffer and fixed with 2 % paraformaldehyde [39]. Data acquisition was performed with acquisition software FACS Diva v6 (BD Biosciences) interfaced with a BD LSR2 flow cytometer (BD Biosciences), and data were analyzed using FLOWJO analysis software v10.2 (Tree Star, Ashland, OR). Gates were assigned according to the appropriate control population.

HIV-1 infection, viral load measurement, and tissue procurement

Four-month-old HSC-engrafted NAIL mice were infected intraperitoneally with 10⁴ TCID₅₀ of macrophage-tropic HIV-1_ADA strain. Mice that received sterile saline injection served as controls. Three HSC-reconstituted NAIL mice were infected with HIV-1, and three saline-injected HSC-reconstituted NAIL mice served as controls. Mice were bled 10 days after the infection to evaluate plasma viral load using the automated COBAS Ampliprep V2.0/Taqman-48 system (Roche, Boston, MA) [40]. Animals were euthanized 25 days post-infection, and tissues were collected for further analysis. Under terminal pentobarbital anesthesia (Fatal Plus), mice were perfused via cardiac puncture with PBS, and the brain was carefully removed. The right hemisphere was separated and placed in 4 % paraformaldehyde (Sigma-Aldrich, St. Louis, MO) in PBS. The left hemisphere was dissected, and the cortices and hippocampi were flash-frozen and stored at −80 °C for biochemical analyses. Tissue viral load in the brain and spleen was measured [39].

Aβ ELISA

Frozen mouse cortices were homogenized in 50 mM Tris-HCL (pH 7.6) containing 150 mM NaCl and a protease inhibitor. Lysates were centrifuged at 20,000 g for 60 min at 4 °C, and the supernatants were collected to assess soluble Aβ₄₂. To assess insoluble Aβ₄₂, pellets were dissolved using 6M guanidine-HCL and centrifuged at room temperature at the same speed and time. Following the manufacturer’s protocol, Aβ₄₂ concentrations in the supernatants were quantified using an ELISA kit (Quantikine^® ELISA, cat. DAB142, R&D Systems, Minneapolis, MN) [41].

Immunohistochemical (IHC) and immunofluorescence (IF) staining

After transcardial perfusion, the right hemispheres of mouse brains were immersed in freshly depolymerized 4 % paraformaldehyde in PBS for 24 h at 4 °C. Brains were then transferred to 70 % ethanol until ready for processing. Brains were processed overnight in an Eperdia™ STP 120 Spin tissue processor (Thermo Fisher Scientific, Waltham, MA) and embedded in paraffin blocks. Sections of 5 μm thick were collected directly onto slides for IHC staining. IHC slides were deparaffinized and antigen retrieval was performed using Trilogy (cat. 922p-07, Sigma-Aldrich)^® reagent as previously described [32]. Brain sections were blocked with 10 % normal goat serum to prevent non-specific staining. Blocked brain sections were incubated overnight at 4 °C with primary mouse monoclonal antibody anti-human leucocyte antigen DR (HLA-DR antibody; cat.NB600-989, Novus Biologicals, Centennial, CO), mouse monoclonal antibody anti-HIV-1 p24 (sc-65918, Santa Cruz, Santa Cruz, CA), rabbit anti-ionized calcium-binding adaptor molecule 1 (IBA1, cat, 013-27691, Fujifilm, Valhalla, NY) or mouse anti-human Aβ (6E10, cat. 803014, Biolegend, San Diego, CA). Following washing, sections were incubated with horseradish peroxidase (HRP)-conjugated anti-mouse/rabbit secondary antibody (cat. GTX83398, GeneTex, Irvine, CA) for 1 h at room temperature. Sections were washed and reacted with hydrogen peroxide substrate and 3,3′-diaminobenzidine (DAB) as chromogen (cat. D4293, Millipore, Sigma, Burlington, MA). Sections were counterstained with hematoxylin (cat. 10,015-074, VWR, Radnor, PA). Slides were imaged using a Zeiss Axioscan 7 whole slide imaging system.

For IF staining, cryopreserved left hemisphere brain tissue was processed as described previously [41]. Briefly, 30 μm thick free-floating brain sections were blocked using 10 % normal goat serum. Following blocking, sections were incubated with mouse monoclonal antibody anti-HLA-DR (Novus, cat. NB600-989) and rabbit antibody anti-Aβ fibril OC (cat. AB2286, Millipore, Sigma) overnight at 4 °C. Following washing, tissue was incubated with Alexa Fluor 568-conjugated goat anti-mouse IgG (cat. A11004, Invitrogen. Carlsbad, CA) and Alexa Fluor 488 goat anti-rabbit IgG (cat. A11008, Invitrogen) secondary antibody for 1 h at room temperature. Images were obtained using a Zeiss Axioscan 7 whole slide imaging system.

Statistical analysis

Data were presented as mean values ± standard error of the mean (SEM). Statistical comparisons of the means were performed with Student’s t-test (GraphPad Software v9.0, San Diego, CA). A p-value of less than 0.05 was considered statistically significant.

Development of single knock in APP(Swe) mutation NOG (NA) mice

The genomic DNA sequences of mouse APP and human APP genes were determined from the NCBI database, and sequences for the Aβ_1–42 peptide encoded by the sequences of exon 16 were analyzed and compared. The two mutations (K670N and M671L) were knocked into exon 16 using CRISPR/Cas9 technology to develop APP knock-in mice containing the Swedish double mutation of familial AD. As additional differences between mouse APP and human APP exon 16 were apparent, we also made nucleotide substitutions that encode for G676R, F681Y, and R684H to humanize the entire mouse exon 16 (Figure 1B). These mice were maintained as homozygous lines.

Generation of NAIL mice

Homozygous NOG/APP^KM670/671NL (NA) mice were crossed with NOD. Cg-Prkdc ^scid Il2rg ^tm1Sug Tg(CMV-IL34)1/Jic (NOG/IL34) mice to generate heterozygous NOG/APP^KM670/671N/IL34 (NAIL) mice. Heterozygous NAIL mice were backcrossed with homozygous NA mice to generate NAIL mice homogeneous for human APP exon 16 and heterogeneous for IL34 gene. qPCR confirming the presence of the human IL-34 gene in NAIL mice (Figure 1C). We maintained the homozygous level of human APP exon 16 in NAIL mice to exclusively express only the human form of Aβ. The presence of human-Aβ in the NAIL mice was confirmed by positive 6E10 antibody staining. The intraneuronal expression of human Aβ in the cortex of NAIL mice was visualized in brain sections from NAIL and background NOG mice. Positive DAB staining was detected in HSC-reconstituted NAIL mice, with or without HIV-1 infection, confirming intraneuronal expression of human Aβ in NAIL mice. No staining was detected in NOG mice, indicating the absence of human Aβ in the mouse strain (Supplementary Figure 1).

HIV-1 infection of NAIL mice and neuropathology

Neonatal (Day 2–3) NAIL mice were engrafted with CD34+ HSC (Figure 1D). By 3-months of age, engrafted NAIL mice were assessed for reconstitution of human immune cells in peripheral blood by flow cytometric analysis. NAIL mice were reconstituted with human CD45+, CD19+ (B cells), CD3+, CD4+, and CD8+ (T cells) as well as CD14+ monocytes (Figure 1E). Reconstituted NAIL mice were placed into infection studies with HIV_ADA by 4 months of age with plasma viral loads determined at days 10 and 25 post-infection and tissues dissected at day 25 for immunohistochemical and immunofluorescence examination (Figure 1D).

Plasma HIV-1 RNA levels showed productive viral infection with an average RNA copy of 1.24 × 10³/ml at day 10 dpi and 4.81 × 10⁴/ml at 25 days after viral infection (Figure 2A). Viral DNA levels in tissues revealed HIV-1 DNA levels approximating 10⁵ copies/10⁶ human CD45+ cells in the spleen and brain by day 25 post-infection, thus indicating the establishment of HIV-1 reservoirs within the human immune cell populations in those tissues (Figure 2B). Spleen sections immunostained for HLA-DR and HIV-1p24 confirmed the presence HLA-DR + reconstitution of human immune cells that reside in the marginal zones of the spleen (Figure 2C). Moreover, numerous foci of HIVp24+ expressing cells were concomitantly found in spleens from infected NAIL mice, but not from uninfected NAIL controls. Brain sections stained for HLA-DR and IBA-1 showed similar expression patterns for IBA-1 and HLA-DR suggesting that IBA-1 reactive microglia are of human origin (Figure 2D). Moreover, substantially greater levels of both HLA-DR+ and IBA-1+ microglia were evident in brains from HIV-1 infected mice compared to uninfected controls. Immunohistochemical staining of brain sections stained for HIVp24 showed expression of the viral proteins in different brain sub-regions including the cortex and hippocampus from HIV-1 infected NAIL mice but not those from uninfected controls (Figure 2E).

Figure 2:

Immune-reconstituted NAIL mice support productive HIV-1 infection. NAIL mice were infected with 10⁴ TCID₅₀ HIV-1_ADA and monitored for 25 days of infection before termination. HSC reconstituted NAIL mice treated with sterile saline served as uninfected controls. (A) Plasma HIV-1 RNA levels were determined on days 10 and 25 post-infection to measure viral load and disease progression. The detection limit (DL), 200 RNA copies/ml. (B) DNA was isolated from brain and spleen cells on day 25 post-infection. HIV-1 DNA levels were determined and normalized to the number of human CD45+ cells to indicate reservoir establishment after infection. Detection limit (DL), 10 DNA copies/10⁶ CD45+ cells. Tissue sections of spleen and brain were prepared from NAIL mice on day 25 post-infection or from uninfected controls. (C) Spleen sections were immunostained for expression of HLA-DR and HIVp24 (red arrowheads) to confirm productive HIV-1 infection in lymphoid tissues. (D) Brain sections were immunostained for HLA-DR and IBA-1 to validate the development of human microglia-like innate immune cells. (E) Immunohistochemical staining images of brain sections from control and HIV-1 infected NAIL mice stained for HIV-p24 protein. Positive HIV-p24 staining confirms productive viral infection in the brain. Scale bars, 100 μm.

Brain sections from human immune-reconstituted NAIL mice stained for Aβ fibrils (OC+) showed numerous Aβ fibrils from HIV-1 infected NAIL mice (Figure 3A). Still, few fibrils were observed in the uninfected control mice (Figure 3B). Co-staining for OC+ Aβ fibrils and expression of HLA-DR showed co-localization of foci of Aβ fibrils with HLA-DR+ cells surrounded by a field of numerous HLA-DR + cells in brain sections from HIV-1 infected NAIL mice, but not from uninfected NAIL controls. Higher-power magnified sections of HIV-1 infected NAIL brains confirmed co-localization of Aβ fibrils with HLA-DR + cells in the brain (Figure 3C). To validate aggregated amyloid pathology, Aβ peptides in lysate fractions from cortical tissues of infected and control NAIL mice were assessed by ELISA. Levels of Aβ₄₂ from insoluble lysate fractions of HIV-1 infected NAIL cortices were 3-fold more significant than those of uninfected NAIL controls. While soluble Aβ₄₂ levels were higher in HIV-1 infected NAIL mice compared to uninfected controls but did not reach statistical significance (Figure 3D). Together, these findings are congruent with increased amyloidogenic pathology in HIV-1 infected NAIL mice compared to uninfected NAIL controls.

Figure 3:

NAIL mice permit simultaneous measures of HIV-1 infection and AD pathobiology. NAIL mice were infected with 10⁴ TCID₅₀ HIV-1_ADA or treated with saline as uninfected controls. Both groups were terminated at day 25 post-infection and sections of brain were evaluated by immunofluorescence. (A) Low (left panel) and high power magnification (right panel) of brain sections from HIV-1 infected NAIL mice stained for Aβ fibrils (OC antibody, green) denoted by gray box or both Aβ fibrils and HLA-DR (human immune cells, red) in the right panel. Scale bars, 100 μm (left panel) and 50 μm (right panel). (B) Brain sections from uninfected control stained for Aβ fibrils (OC antibody, green). Scale bar, 100 μm. (C) High power magnification of brain sections from HIV-1 infected NAIL mice stained for Aβ fibrils (OC antibody, green) and HLA-DR showing co-localization of human immune cells with Aβ-fibrils. Scale bar, 50 μm. (D) Cortical brain tissues from uninfected and day 25 HIV-1 infected NAIL mice were dissected, lysed, and fractionated by differential centrifugation. Soluble (Tris-HCl) and insoluble (Gu-HCl) lysate fractions were evaluated for Aβ₄₂ levels by ELISA. Means ± SEM and significant differences were determined for three mice per group whereby *p≤0.05.