Simple and highly specific targeting of resident microglia with adeno-associated virus

Abstract

Microglia, as the immune cells of the central nervous system (CNS), play dynamic roles in both healthy and diseased conditions. The ability to genetically target microglia using viruses is crucial for understanding their functions and advancing microglia-based treatments. We here show that resident microglia can be simply and specifically targeted using adeno-associated virus (AAV) vectors containing a 466-bp DNA fragment from the human IBA1 (hIBA1) promoter. This targeting approach is applicable to both resting and reactive microglia. When combining the short hIBA1 promoter with the target sequence of miR124, up to 98% of transduced cells are identified as microglia. Such a simple and highly specific microglia-targeting strategy may be further optimized for research and therapeutics.

Keywords: Genetics; Immunology; Molecular Genetics.

Graphical abstract

Figure 1

In vivo screens for microglia-targeting lentivirus See also Data S1. (A) Schematic diagram of the experimental procedure. One week post MCAO-induced stroke, lentiviruses with various promoter-driven GFP were injected into the striata and examined after another week (wk). (B) Representative confocal images showing marker expression for the indicated lentiviruses. Scale bars, 50 μm. (C) Quantifications showing high microglia-specificity of GFP expression in mice injected with lenti-hIBA1-GFP (mean ± SEM; n = 3 mice per group). (D) Representative confocal images showing marker expression for the indicated lentiviruses. Scale bars, 50 μm. (E) Quantifications showing comparable specificity of lentivirus packaged with either VSV-G or LCMV-WE envelope (mean ± SEM; n = 3 mice per group).

Figure 2

In vivo screens for microglia-targeting AAVs (A) Schematic diagram of the experimental procedure. One week post L-NIO-induced stroke, scAAVs with various capsids were injected into the striata and examined after another week (wk). (B) Representative confocal images showing marker expression for the indicated scAAVs. Scale bars, 50 μm. (C) Quantifications showing microglia-specificity of GFP expression for the indicated scAAVs (mean ± SEM; n = 3 mice per group). (D) Quantifications showing microglia transduction efficiency for the indicated scAAVs (mean ± SEM; n = 3 mice per group). (E) Schematic diagram of the experimental procedure. One week post L-NIO-induced stroke, scAAV5 or scAAV8 was injected into the striata and examined after another 4 weeks (wk). (F) Representative confocal images showing marker expression for the indicated scAAVs. Scale bars, 50 μm. (G) Quantifications showing the specificity and transduction efficiency for microglia (mean ± SEM; n = 3 mice per group).

Figure 3

In vivo screens for the minimal microglia-targeting promoter See also Figure S1 and Data S1. (A) Schematic diagram of the experimental procedure. ssAAV5 viruses with different promoters were injected into sham mice or mice with L-NIO-induced stroke. Brains were analyzed one week later. (B) Diagram of the examined ssAAVs with different lengths of hIBA1 promoter. (C) Representative confocal images showing marker expression for the indicated ssAAV5s. Scale bars, 50 μm. (D) Quantifications showing microglia-specificity of GFP expression for the indicated ssAAV5s (mean ± SEM; n = 3 mice per group). (E) Quantifications showing microglia transduction efficiency for the indicated ssAAV5s (mean ± SEM; n = 3 mice per group).

Figure 4

Increased long-term specificity and efficiency for the minimal hIBA1a promoter when packaged in scAAV (A) Schematic diagram of the experimental procedure. scAAV5 virus with the hIBA1a promoter was injected into sham mice or mice with L-NIO-induced stroke. Brains were analyzed 1 week or 4 weeks later. (B) Representative confocal images showing marker expression for the indicated conditions. Scale bars, 50 μm. (C) Quantifications showing microglia-specificity of GFP expression for the indicated conditions (mean ± SEM; n = 3 mice per group). (D) Quantifications showing microglia transduction efficiency for the indicated conditions (mean ± SEM; n = 3 mice per group).

Figure 5

miR124T confers high specificity of ssAAV under the hIBA1a promoter (A) Schematic diagram of the experimental procedure. The miR124T sequence was inserted into the 3′ end of the GFP gene. ssAAV5 virus was then injected into the striatum of adult wild-type mouse and analyzed 4 weeks later. (B) Representative confocal images showing marker expression in the injection area and an area away from the injection site. Neurons are marked by NeuN staining. Scale bars, 50 μm. (C) Quantifications showing high microglia-specificity of GFP expression (mean ± SEM; n = 3 mice per group). (D) Quantifications showing high microglia transduction efficiency (mean ± SEM; n = 3 mice per group).

Figure 6

High microglia-specificity in diverse brain regions See also Figure S2. (A) Representative confocal images showing marker expression in the injection site and an area away from the injection site. Mouse brains were examined 4 weeks post injection of the ssAAV5-hIBA1a-GFP-miR124T virus. GFP signals were from autofluorescence. Scale bars, 50 μm. (B) Representative confocal images showing marker expression in the hippocampus. GFP signals were from autofluorescence. The arrowheads show examples of GFP⁺IBA1⁺ or GFP⁺PU.1⁺ microglia, while the arrows indicate GFP are not detectable in NeuN⁺, GFAP⁺, or OLIG2⁺ cells. The orthogonal views are also shown. Scale bars, 50 μm. (C) Quantifications showing high microglia-specificity of GFP expression in the hippocampus (mean ± SEM; n = 4 mice per group; n.d., not detected).