Recombinant AAV vectors and uses thereof

Abstract

The invention provides a recombinant adeno-associated virus (rAAV) capsid protein which comprises an amino acid sequence of 11 continuous amino acids X1X2X3X4X5X6X7X8X9X10Q, an rAAV vector containing the sequence and application of the rAAV vector.

Description

Technical Field

[0001] The present disclosure relates to the field of adeno-associated virus (AAV) vectors, particularly recombinant AAV (rAAV) capsid proteins, rAAV vectors containing the rAAV capsid proteins, and their uses. Background Art

[0002] Microglia account for approximately 10% of the total number of cells in the central nervous system (CNS). Microglia were initially considered debris scavengers and are now considered the main regulators of the CNS under normal and pathological conditions. Microglia conduct active surveillance and initiate rapid innate and adaptive immune responses when encountering immune attacks. In addition to their functions in immunity, recent studies have revealed the multifaceted roles of microglia in controlling neural circuit development and plasticity. Emerging evidence indicates that microglial dysfunction is a key factor in the progression of CNS aging and CNS diseases, including neurological diseases and brain cancers. Many clinical studies have identified risk-related alterations in genes highly expressed by microglia, highlighting the involvement of microglia in the progression of CNS diseases and the potential for targeted microglial treatment interventions.

[0003] A key challenge in studying microglia is the inability to effectively label and manipulate them. Current methods rely heavily on generating germline transgenic mouse models to introduce transgenes or genetic modifications into microglia, which is time-consuming, laborious, and often inefficient. In addition, the involvement of germline transgenes hinders their widespread application in animal models with low reproductive rates and long generation times (e.g., non-human primates) and cannot be used as a therapeutic tool for humans.

[0004] Recombinant viral vectors are an attractive alternative for manipulating microglia and have broad prospects for microglial gene therapy. In particular, due to the lack of obvious pathogenicity, recombinant adeno-associated virus (rAAV) is the most commonly used viral vector in current basic research and gene therapy. However, although rAAV can transduce a wide range of mammalian cell types, the transduction of rAAV into microglia is still very poor. rAAV packaged with existing AAV capsids does not achieve high transduction rates and sufficient transgene expression levels in microglia, especially in vivo.

[0005] On the other hand, viral transduction of microglia (and macrophages in general) also faces the potential problem of inducing immune activation. For example, recombinant adenoviruses effectively transduce macrophages but at the same time cause the transduced cells to produce an immune response.

[0006] Therefore, there is still a need to develop a system to achieve high transduction rates and sufficient transgene expression levels in microglia without triggering immune reactivation of the transduced cells. SUMMARY OF THE INVENTION

[0007] To overcome at least one of the above technical problems, the present disclosure provides novel AAV capsids that mediate efficient gene delivery to microglia and their applications.

[0008] The present disclosure provides recombinant adeno-associated virus (rAAV) capsid proteins that have a 7-amino acid peptide insertion () compared to parental AAV capsid proteins. When the rAAV capsid proteins provided herein are present in AAV vectors / virus particles, they can improve the transduction efficiency of microglia both in vivo and in vitro compared to AAV virus particles without the peptide insertion. Also provided are recombinant AAV vectors / virus particles and pharmaceutical compositions comprising the rAAV capsid proteins provided therein; and methods of using these rAAV capsid proteins and vectors / virus particles in research and clinical practice, for example, delivering polynucleotide sequences to microglia to treat microglia-related diseases.

[0009] According to one aspect of the present disclosure, there is provided a recombinant adeno-associated virus (rAAV) capsid protein comprising the amino acid sequence of 11 consecutive amino acids X1X2X3X4X5X6X7X8X9X 10 Q, wherein X1 is selected from Ala or Leu; X2 is selected from Gln, Met, Thr, Val or Pro; X3 is selected from Trp, Thr, Glu, Pro, Leu, Ala or Gln; X4 is selected from Pro, Thr, Met, Ser, Arg or Ala; X5 is selected from Pro, Ser, Val, Asp or Phe; X6 is selected from Lys or Pro; X7 is selected from Thr or Arg; X8 is selected from Thr, Glu or Pro; X9 is selected from Ser, Pro or Ala; and X 10 is selected from Ala or Asp.

[0010] In some embodiments, the rAAV capsid protein comprises the amino acid sequence of 11 consecutive amino acids X1X2X3X4X5X6X7X8X9X 10 Q, wherein X6 is Lys; X7 is Thr; X8 is Thr; X9 is Ser; and / or X 10It is Ala. In some embodiments, the rAAV capsid protein comprises the amino acid sequence of 11 consecutive amino acids X1X2X3X4X5KTTSAQ. In some other embodiments, X1 is selected from Ala or Leu. In some other embodiments, X2 is selected from Gln, Met, Thr or Val. In some other embodiments, X3 is selected from Trp, Thr, Glu, Pro or Leu. In some other embodiments, X4 is selected from Pro, Thr, Met or Ser. In some other embodiments, X5 is selected from Ser, Val, Asp or Pro. In some specific embodiments, X1 is Ala, X2 is Gln, X3 is Trp, X4 is Pro, X5 is Pro. In some specific embodiments, X1 is Leu, X2 is Met, X3 is Thr, X4 is Pro, X5 is Pro. In some specific embodiments, X1 is Ala, X2 is Thr; X3 is Glu, X4 is Pro, X5 is Pro. In some specific embodiments, X1 is Ala, X2 is Gln, X3 is Pro, X4 is Thr, X5 is Ser. In some specific embodiments, X1 is Ala, X2 is Gln, X3 is Leu, X4 is Met, X5 is Val. In some specific embodiments, X1 is Ala, X2 is Gln, X3 is Trp, X4 is Thr, X5 is Asp. In some specific embodiments, X1 is Ala, X2 is Val, X3 is Leu, X4 is Ser, X5 is Pro.

[0011] In some specific embodiments, the rAAV capsid protein comprises the amino acid sequence of AQWPPKTTSAQ (SEQ ID NO.:1). In some specific embodiments, the rAAV capsid protein comprises the amino acid sequence of LMTPPKTTSAQ (SEQ ID NO.:2). In some specific embodiments, the rAAV capsid protein comprises the amino acid sequence of ATEPPKTTSAQ (SEQ ID NO.:3). In some specific embodiments, the rAAV capsid protein comprises the amino acid sequence of AQPTSKTTSAQ (SEQ ID NO.:71). In some specific embodiments, the rAAV capsid protein comprises the amino acid sequence of AQLMVKTTSAQ (SEQ ID NO.:72). In some specific embodiments, the rAAV capsid protein comprises the amino acid sequence of AQWTDKTTSAQ (SEQ ID NO.:73). In some specific embodiments, the rAAV capsid protein comprises the amino acid sequence of AVLSPKTTSAQ (SEQ ID NO.:74).

[0012] In some embodiments, the rAAV capsid protein comprises the amino acid sequence of X1X2X3X4X5X6X7X8X9X 10 Q, wherein X1 is Ala; X6 is Pro; and / or X7 is Arg. In some embodiments, the rAAV capsid protein comprises the amino acid sequence of AX2X3X4X5PRX8X9X 10 Q. In some other embodiments, X2 is selected from Gln or Pro. In some other embodiments, X3 is selected from Thr, Ala or Gln. In some other embodiments, X4 is selected from Arg or Ala. In some other embodiments, X5 is selected from Pro or Phe. In some other embodiments, X8 is selected from Glu or Pro. In some other embodiments, X9 is selected from Pro or Ala. In some other embodiments, X 10 is selected from Ala or Asp. In some specific embodiments, X2 is Gln, X3 is Gln, X4 is Arg, X5 is Pro, X8 is Glu, X9 is Pro, X 10 is Ala. In some specific embodiments, X2 is Gln, X3 is Gln, X4 is Arg, X5 is Pro, X8 is Pro, X9 is Ala, X 10 is Asp. In some specific embodiments, X2 is Gln, X3 is Thr, X4 is Ala, X5 is Phe, X8 is Glu, X9 is Pro, X 10 is Ala. In some specific embodiments, X2 is Pro, X3 is Ala, X4 is Arg, X5 is Pro, X8 is Glu, X9 is Pro, X 10 is Ala.

[0013] In some specific embodiments, the rAAV capsid protein comprises the amino acid sequence of AQQRPPREPAQ (SEQ ID NO.:4). In some specific embodiments, the rAAV capsid protein comprises the amino acid sequence of AQQRPPRPADQ (SEQ ID NO.:5). In some specific embodiments, the rAAV capsid protein comprises the amino acid sequence of AQTAFPREPAQ (SEQ ID NO.:75). In some specific embodiments, the rAAV capsid protein comprises the amino acid sequence of APARPPREPAQ (SEQ ID NO.:76).

[0014] According to another aspect of the present disclosure, there is provided a polynucleotide sequence encoding the rAAV capsid protein provided by the present disclosure.

[0015] According to another aspect of the present disclosure, there is provided a recombinant adeno-associated virus (rAAV) vector comprising a capsid protein provided by the present disclosure. In some embodiments, the rAAV vector further comprises a heterologous polynucleotide sequence. In some embodiments, the heterologous polynucleotide sequence encodes a heterologous polypeptide, non-coding RNA, or CRISPR agent.

[0016] In some embodiments, the CRISPR agent comprises DNA-targeting RNA, such as crRNA-like RNA, tracrRNA-like RNA, single-guide RNA, etc. In some embodiments, the heterologous polynucleotide sequence encodes a protein, such as an antibody, a membrane protein (such as a receptor), a chaperone protein, or a ubiquitin ligase. In some embodiments, the heterologous polynucleotide sequence encodes miRNA, siRNA, piRNA, lncRNA, or guide RNA.

[0017] According to another aspect of the present disclosure, there is provided a pharmaceutical composition comprising the rAAV vector / virus particle provided by the present disclosure, and a pharmaceutically acceptable carrier.

[0018] According to another aspect of the present disclosure, there is provided a method for delivering the rAAV vector provided by the present disclosure to a target cell, which comprises contacting the target cell with the rAAV vector / virus particle. In some embodiments, the target cell is a microglial cell in vitro or in vivo. In some embodiments, the contacting is carried out in the presence of an inhibitor of topoisomerase or proteasome. In certain embodiments, the contacting is carried out in the presence of a topoisomerase inhibitor and / or a DNA damage inducer. The topoisomerase inhibitor can be selected from doxorubicin (a DNA topoisomerase II inhibitor), bortezomib (a proteasome inhibitor), etoposide (a DNA topoisomerase II inhibitor), teniposide (a DNA topoisomerase II inhibitor), vanillin (a non-homologous end joining inhibitor), etc. The DNA damage inducer can be bleomycin, etc.

[0019] According to another aspect of the present disclosure, there is provided a host cell comprising the polynucleotide sequence encoding the rAAV capsid protein provided by the present disclosure.

[0020] According to another aspect of the present disclosure, there is provided a method for treating a neurological disorder, which comprises administering to a subject in need thereof a therapeutically effective amount of the pharmaceutical composition. In some embodiments, the neurological disorder can be a disease related to microglial cells. In some embodiments, the neurological disorder can include Alzheimer's disease, Parkinson's disease, Huntington's disease, and amyotrophic lateral sclerosis, leukoencephalopathy, glioma, and atypical teratoid / rhabdoid tumor.

[0021] The development of engineered AAV capsids capable of mediating efficient microglial transduction provides much-needed viral tools for studying microglial biology. This disclosure demonstrates that the rAAV provided herein enables sufficient transgene expression in microglia for labeling, monitoring, and manipulation. The newly evolved AAV capsids herein can facilitate the application of various genetically encoded tools (such as fluorescent sensors for signaling molecules, optogenetic and chemogenetic effectors) and gene editing methods in microglia in vitro and in vivo. Combining the rAAV provided herein with other rAAVs targeting other cell types in the CNS also represents a promising strategy for studying the interactions between microglia and different cell types in the same animal. Recent single-cell transcriptomics studies have revealed unexpectedly large regional heterogeneity of microglia in the brain. Thus, the rAAV of this disclosure may be an ideal tool for in vivo brain region-specific microglial manipulation to study the role of microglia in controlling neural circuits in different brain regions.

[0022] Genetic studies in human patients have identified many drug targets and signaling pathways for CNS diseases that are highly enriched in microglia. Introducing disease-related mutations into mouse models leads to microglial dysfunction and induces pathological and behavioral phenotypes similar to those in human patients, highlighting the great potential of microglia-based gene therapy. As a key prerequisite for gene therapy, the effective transfer of therapeutic agents to target cells requires a suitable vector. The lack of safe, effective, and clinically relevant delivery methods has hindered the development of microglial gene therapy. The successful evolution of the AAV capsids of this disclosure demonstrates the possibility of using rAAV for efficient microglial transduction and lays the foundation for future optimization of AAV capsids for microglia-based gene therapy.

[0023] In other embodiments, the AAV vector / viral particle comprising the variant capsid protein described above may include any of the embodiments described above or subsequently disclosed. In fact, it should be understood that certain features disclosed herein, described in the context of separate embodiments for clarity, may also be provided in combination in a single embodiment. Conversely, the various features disclosed herein, described in the context of a single embodiment for brevity, may also be provided separately or in any suitable sub-combination. All combinations of embodiments related to the present invention are explicitly included in and disclosed in the present invention as if each and every combination were individually and explicitly disclosed. In addition, all sub-combinations of the various embodiments and their elements are specifically included in and disclosed in the present invention as if each and every such sub-combination were individually and explicitly disclosed herein. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1Directed evolution generates AAV-MGs that mediate efficient microglial transduction in vitro and in vivo. (A) Schematic of the in vitro screening process. A heptamer was randomly inserted between the 588th and 589th amino acids of the AAV9 VP1 protein. The library was screened in cultured mouse microglia for two rounds. (B) Distribution of AAV9 capsid variants recovered from cultured mouse microglia, sorted in descending order of enrichment score. The pie chart shows the normalized frequency of AAV-cMG.WPP in the total recovered sequences. Cyan: AAV-cMG.WPP. (C) Schematic of the in vivo selection process. Semi-random mutations were introduced into the heptamer inserted in AAV-cMG.WPP and its four neighboring amino acids. The library was screened in the striatum and midbrain of Cx3cr1 CreER mice. Capsid variants of the Cre-recombined AAV genome were selectively recovered from the brain. (D) Distribution of AAV-cMG.WPP variants recovered from the brains of Cx3cr1 CreER mice, sorted in descending order of enrichment score. The pie chart shows the normalized frequencies of AAV-MG1.1 and AAV-MG1.2 in the total recovered sequences. Magenta: AAV-MG1.1, Green: AAV-MG1.2, Cyan: AAV-cMG.WPP. (E) Representative images of mScarlet expression patterns in the orbitofrontal cortex (OFC, top) and striatum (bottom) of Cx3cr1 CreER mice injected with AAV-cMG.WPP-SFFV-DIO-mScarlet, AAV-MG1.1-SFFV-DIO-mScarlet, or AAV-MG1.2-SFFV-DIO-mScarlet. AAVs were injected with or without doxycycline administered posterior to the orbit. Detailed quantification results are shown in Figure 5 A. (F) Representative images of co-localization of mScarlet and Iba CreER immunostaining signals in the OFC and striatum of Cx3cr1 + mice injected with AAV-MG1.1-SFFV-DIO-mScarlet. Detailed quantification results are shown in Figure 5 D. Scale bars, 250 μm (E), 100 μm (F).

[0025] Figure 2.AAV-cMG mediates efficient microglial transduction in vitro. (A) Schematic diagram of the selection process. Seven amino acids were randomly inserted between amino acids 588 and 589 of the AAV9VP1 protein. The library was screened for two rounds in cultured mouse microglia. (B) Distribution of AAV9 capsid variants recovered from cultured mouse microglia, sorted by descending enrichment score. The pie chart shows the normalized frequency of AAV-cMG.QRP in the total recovered sequences. (C) Representative images of cultured mouse microglia transduced with mScarlet reporter AAV packaged with AAV-MG.QRP. (D) Schematic diagram of the selection process for AAV-MG.QRP variants. The right panel shows the distribution of AAV-MG.QRP variants recovered from cultured mouse microglia, sorted by descending enrichment score. The pie chart shows the normalized frequency of AAV-cMG in the total recovered sequences. (E) Representative images of cultured mouse microglia transduced with mScarlet reporter AAVs packaged with different capsids. (F) Representative images of mScarlet and Iba in cultured mouse microglia transduced with AAV-cMG-SFFV-mScarlet + Representative immunofluorescence images of colocalization of immune signals. (G) mScarlet expression in cultured mouse microglia transduced with mScarlet reporter AAVs packaged with different capsids + Quantification of percentage and mean fluorescence intensity (n=6 replicates per group in 3pt; 5 replicates per group in 5pt; bar graphs represent the mean of each group; one-way analysis of variance (ANOVA) with Dunnett's post hoc test). (H) Representative images of cultured microglia transduced with AAV-cMG-SFFV-mScarlet with (right) or without (left) doxorubicin are shown. (I) Representative images of cultured mouse microglia transduced with AAV-cMG-SFFV-mScarlet with or without doxorubicin + Quantification of percentage and mean fluorescence intensity (n=6 replicates per group; bar graphs represent the mean of each group; one-way ANOVA with Dunnett's post hoc test). Scale bars, 200 μm (C, E, H), 50 μm (F).

[0026] Figure 3 . Transduction of cultured mouse microglia by AAV-cMG.WPP. (A) Representative images of cultured mouse microglia transduced with mScarlet reporter rAAVs packaged with different capsids. (B) Representative images of mScarlet and Iba in cultured mouse microglia transduced with mScarlet reporter rAAVs packaged with AAV-cMG.WPP. +Representative immunofluorescence images of co - localization of immune signals. (C) mScarlet of cultured mouse microglia transduced with mScarlet reporter rAAV packaged with different capsids + Quantification of percentage and mean fluorescence intensity (n = 4 replicates per group; bars represent the mean of each group; one - way ANOVA with Dunnett's post - hoc test). (D) Principal component analysis of the transcriptomes of cultured mouse microglia in four treatment groups: control non - transduced (steady state), LPS - treated (responsive state), interleukin - 4 - treated (IL4; alternative activation state), and AAV - cMG.WPP - transduced group (n = 3 replicates per group). (E) Heatmap showing the expression of marker genes for microglial responsive and alternative activation states in different treatment groups in (D). Scale bars, 100 μm (A), 50 μm (B).

[0027] Figure 4 . AAV - MG mediates efficient microglial transduction in vivo. (A) Representative images of mScarlet expression patterns in the midbrain of Cx3cr1 mice injected with AAV - cMG.WPP - SFFV - DIO - mScarlet (left), AAV - MG1.1 - SFFV - DIO - mScarlet (middle), or AAV - MG1.2 - SFFV - DIO - mScarlet (right) CreER Representative images of mScarlet expression patterns in the hippocampus and thalamus of Cx3cr1 mice injected with AAV - MG1.1 - SFFV - DIO - mScarlet (left) or AAV - MG1.2 - SFFV - DIO - mScarlet (right) CreER Scale bar, 500 μm.

[0028] Figure 5 . Quantification of AAV - MG - mediated microglial transduction in vivo. (A) Cx3cr1 mice injected with AAV - cMG.WPP - SFFV - DIO - mScarlet, AAV - MG1.1 - SFFV - DIO - mScarlet, or AAV - MG1.2 - SFFV - DIO - mScarlet CreERMicroglial cell counts of mScarlet-labeled in the mouse OFC, striatum, and midbrain (n = 3 mice per group; two-way ANOVA with Tukey's post hoc test; P < 0.0001 for cMG.WPP vs. MG1.1 in the OFC, striatum, and midbrain; P < 0.0001 for cMG.WPP vs. MG1.2 in the OFC, striatum, and midbrain). The x-axis represents the distance (μm) of the brain slice from the virus injection site. (B) Cx3cr1 injected with AAV-MG1.1-SFFV-DIO-mScarlet (left) or AAV-MG1.2-SFFV-DIO-mScarlet (right) CreER Fluorescence intensity distribution of mScarlet-labeled microglial cells in the mouse striatum (n = 3 mice per group). AAV was injected with or without doxorubicin (Doxo) administered posterior to the orbit. (C) Cx3cr1 injected with AAV-MG1.1-SFFV-DIO-mScarlet CreER mScarlet and Iba in the mouse midbrain + Representative images of co-localization of immunofluorescence signals. (D) Quantification of the percentage of mScarlet- and Iba-double positive microglial cells among the total mScarlet-positive microglial cells (n = 3 mice per group). (E) Cx3cr1 injected with AAV5-SFFV-DIO-mScarlet (left), AAV6TM-SFFV-DIO-mScarlet (middle), or AAV9-SFFV-DIO-mScarlet (right) CreER Representative images of the mScarlet expression pattern in the mouse striatum. Scale bars, 100 μm (C), 100 μm (E).

[0029] Figure 6.AAV-MG transduction of microglia in vivo does not induce microglia activation. (A) UMAP plot of 197 microglia isolated from 3 mice (90 non-transduced, 11 transduced with AAV-MG1.1, 96 transduced with AAV-MG1.2). The inset shows the log-normalized expression levels of the mScarlet transcript. (B) Violin plots showing the expression levels of homeostatic marker genes (Cx3cr1, Tmem119, P2ry12, and Csf1r) and reactive marker genes (Cd74, Tlr2, Cebpb, and Spp1). (C) UMAP plot recalculated from a reference microglia dataset (4500 microglia from control mice and 9832 microglia from LPS-treated mice sequenced using the 10x Genomics platform) and a query dataset (sequenced using the Smart-seq2 protocol). The inset summarizes the states of the microglia sequenced by Smart-seq2, as predicted by Seurat label transfer.

[0030] Figure 7 .Characterize the single-cell transcriptomes of homeostatic and reactive microglia using reference single-cell RNA sequencing datasets. (A) Figure 6 Quantification of gene counts, mitochondrial RNA percentage, and ribosomal RNA percentage in the Smart-seq2 dataset in A. Red dots indicate sequenced microglia that did not pass quality control and were removed from subsequent analyses. A total of 197 single microglia with positive quality control (QC) were obtained (90 non-transduced, 11 transduced with AAV-MG1.1, 96 transduced with AAV-MG1.2), with an average gene detection rate of 3961 genes per cell. (B) UMAP plot of 14332 microglia in the reference microglia dataset [4500 from control mice and 9832 from LPS-treated mice (0.83 mg / kg, intraperitoneal injection), sequenced using the 10x Genomics platform]. (C) Violin plots showing the expression levels of homeostatic marker genes (Cx3cr1, Tmem119, P2ry12, and Csf1r) and reactive marker genes (Cd74, Tlr2, Cebpb, and Spp1) in the reference microglia dataset. (D) Quantification of gene expression changes in microglia in the homeostatic and reactive states in the reference microglia dataset. Red dots indicate reactive marker genes.

[0031] Figure 8 .AAV-MG can achieve microglial Ca 2+In vivo two-photon imaging of signal and ATP transmission. (A) Schematic diagram of the experimental procedure for in vivo two-photon imaging of calcium signal and ATP transmission. (B) Images of GCaMP8 expression in microglia and heatmaps of GCaMP8 fluorescence signals 3 hours after injection of LPS (left) or saline (right). (C) Quantification of GCaMP8 fluorescence signals in microglia somata after intraperitoneal injection of LPS (red) or saline (black) (n = 14 cells in the LPS group, n = 10 cells in the saline group; two-way ANOVA, #: Tukey's post hoc vs. 0 hours, *: Sidak's post hoc vs. saline). (D) Images of GRAB ATP1.0 expression in microglia and heatmaps of GRAB ATP1.0 fluorescence signals 3 hours after injection of LPS (left) or saline (right). (E) Quantification of GRAB ATP1.0 fluorescence signals in microglia somata after intraperitoneal injection of LPS (red) or saline (black) (n = 14 cells in the LPS group, n = 10 cells in the saline group; two-way ANOVA, #: Tukey post hoc analysis vs. 0 hours, *: Sidak post hoc analysis vs. saline). Error bars represent s.e.m. Scale bar, 100 μm (B, D).

[0032] Figure 9 . In vivo two-photon imaging of extracellular ATP changes in microglia after acute laser ablation of AAV-MG1.2. (A) Quantification of GRAB ATP1.0 fluorescence signals in microglia somata of control mice (black) or mice receiving laser ablation (red) (n = 10 cells in the control group, n = 16 cells in the laser ablation group; two-way ANOVA). (B) Expression images of GRAB ATP1.0 in microglia and heatmaps of GRAB ATP1.0 fluorescence signals at 10 min (left), 20 min (middle), and 40 min (right) during imaging. For the laser ablation group, the laser was applied to the center of the field of view at the start of the imaging process (0 min). Error bars represent s.e.m. Scale bar, 100 μm.

[0033] Figure 10 . In vivo efficient microglial genome editing mediated by AAV-MG. (A) Cx3cr1 injected with AAV-MG1.1-sgRNA-LacZ (left) or AAV-MG1.1-sgRNA-Tmem119 (right) CreER: Representative immunofluorescence images of the striatum of Rosa26-LSL-Cas9 mice. Brain sections were immunostained with Tmem119. The upper right panel shows an enlarged view of the boxed area in the left panel. The lower right panel shows the percentage of Tmem119-positive pixels in a 1.5 mm × 1.5 mm area of the dorsal striatum (n = 3 mice per group; two-way ANOVA). The x-axis represents the distance (μm) of the brain section from the virus injection site. (B) Cx3cr1 injected with AAV-MG1.2-sgRNA-LacZ (left) or AAV-MG1.2-sgRNA-Cd68 (right) CreER : Representative immunofluorescence images of the striatum of Rosa26-LSL-Cas9 mice. Brain sections were immunostained with Cd68. The upper right panel shows an enlarged view of the boxed area in the left panel. The lower right panel shows the number of Cd68-positive cells in a 1 mm × 1 mm area of the dorsal striatum (n = 3 mice per group; two-way ANOVA). The x-axis represents the distance (μm) of the brain section from the virus injection site. (C) Schematic diagram of the experimental procedure for in vivo two-photon imaging of microglial responses to laser ablation. (D) Cx3cr1 injected with AAV-MG1.2-CMV-SaCas9 (black) or AAV-MG1.2-CMV-SaCas9-U6-sgRNA-P2ry12 (red) GFP Quantification of microglial process extension at the laser ablation site in Cx3cr1 mice (n = 4 fields of view for 3 mice injected with AAV-MG1.2-CMV-SaCas9, n = 6 fields of view for 3 mice injected with AAV-MG1.2-CMV-SaCas9-U6-sgRNA-P2ry12; two-way ANOVA). (E) Cx3cr1 mice injected with AAV-MG1.2-CMV-SaCas9 (left) or AAV-MG1.2-CMV-SaCas9-U6-sgRNA-P2ry12 (right) GFP Representative images of the recruitment process of microglia to the two-photon laser ablation site in Cx3cr1 mice. Error bars represent s.e.m. Scale bars, 200 μm (A left, B left), 100 μm (A right, B right), 20 μm (E).

[0034] Figure 11 . AAV-MG-mediated in vivo microglia-specific gene knockout. (A) Cx3cr1 injected with AAV-MG1.1-sgRNA-LacZ (upper) or AAV-MG1.1-sgRNA-Tmem119 (lower) CreER: Representative immunofluorescence images of the striatum of Rosa26-LSL-Cas9 mice. Brain sections were immunostained with Tmem119. (B) Cx3cr1 injected with AAV-MG1.2-sgRNA-LacZ (top) or AAV-MG1.2-sgRNA-Cd68 (bottom) CreER : Representative immunofluorescence images of the striatum of Rosa26-LSL-Cas9 mice. Brain sections were immunostained with Cd68. (C) Cx3cr1 after injection of AAV-MG1.2-CMV-SaCas9 (top) or AAV-MG1.2-CMV-SaCas9-U6-sgRNA-P2ry12 (bottom) GFP Representative images of the S1 cortex of mice. Brain sections were immunostained with P2ry12. (D) Cx3cr1 injected with AAV-MG1.2-CMV-SaCas9 (black) or AAV-MG1.2-CMV-SaCas9-sgRNA-P2ry12 (red) GFP Quantification of the percentage of P2ry12-positive pixels in a 1 mm × 1 mm area of the S1 cortex of mice (n = 3 mice in the P2ry12 knockout group; n = 2 mice in the control group; two-way ANOVA). Error bars represent s.e.m. Scale bars, 500 μm (A, B), 100 μm (C).

[0035] Figure 12 . Cx3cr1 injected with (A) AAV-MG.TAF-SFFV-DIO-mScarlet, (B) AAV-MG.APA-SFFV-DIO-mScarlet, (C) AAV-MG.PTS-SFFV-DIO-mScarlet, (D) AAV-MG.LMV-SFFV-DIO-mScarlet, (E) AAV-MG.WTD-SFFV-DIO-mScarlet, or (F) AAV-MG.VLS-SFFV-DIO-mScarlet CreER Representative images of the mScarlet expression pattern in the striatum of mice. Scale bar, 500 μm. Detailed implementation

[0036] Before describing the methods and compositions, it should be understood that the present invention is not limited to the specific methods or compositions described, which can vary. It should also be understood that the terms used herein are for the purpose of describing specific embodiments only and are not intended to be limiting, as the scope of the present invention will be limited only by the appended claims.

[0037] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Although any methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present invention, some potential and preferred methods and materials are now described. All publications mentioned herein are incorporated by reference herein to disclose and describe the methods and / or materials related to the cited publications. It should be understood that to the extent there is any conflict, the present disclosure supersedes any disclosure of the incorporated publications.

[0038] It is noted that as used herein and in the appended claims, the singular forms "a", "an", and "the" include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to "a recombinant AAV viral particle" includes a plurality of such viral particles, reference to "microglia" includes reference to one or more microglia known to those of skill in the art and their equivalents, and so forth.

[0039] Definition

[0040] Unless otherwise defined, all scientific and technical terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this technology belongs.

[0041] Adeno-associated virus (AAV) is a member of the Parvoviridae family and belongs to the Dependovirus genus. AAV is a non-pathogenic parvovirus composed of a single-stranded DNA genome of approximately 4.7 kb within an unenveloped icosahedral capsid. The genome contains three open reading frames (ORFs) flanked by inverted terminal repeats (ITRs), which serve as the origin of viral replication and packaging signals. The rep ORF encodes four non-structural proteins that play roles in viral replication, transcriptional regulation, site-specific integration, and viral particle assembly. The cap ORF encodes three structural proteins (VP 1-3), which assemble to form a 60-mer viral capsid. Finally, the ORF generates an assembly activation protein (AAP) as an alternative reading frame within the cap gene, and AAP is a viral protein that localizes AAV capsid proteins to the nucleolus and plays a role in capsid assembly. Based on the crystal structure of AAV, the VP amino acids involved in forming the icosahedral five-fold, three-fold, and two-fold symmetry interfaces have been visualized. Surface loops on the three-fold symmetry axis are thought to be involved in binding to host cell receptors and have been the target of mutagenesis studies.

[0042] There are several naturally occurring (“wild-type”) serotypes of AAV and over 100 known variants, each with a different amino acid sequence, particularly in the hypervariable regions of the capsid proteins, and thus different gene delivery properties. AAV is not associated with any human diseases, which makes recombinant AAV attractive for clinical applications.

[0043] Unless otherwise indicated, the term “adeno-associated virus” or “AAV” refers to all subtypes or serotypes, as well as replication-competent and recombinant forms. The term “AAV” includes, but is not limited to, AAV type 1 (AAV-1 or AAV1), AAV type 2 (AAV-2 or AAV2), AAV type 3 (AAV-3A or AAV3A), AAV type 3B (AAV-3B or AAV3B), AAV type 4 (AAV-4 or AAV4), AAV type 5 (AAV-5 or AAV5), AAV type 6 (AAV-6 or AAV6), AAV type 7 (AAV-7 or AAV7), AAV type 8 (AAV-8 or AAV8), AAV type 9 (AAV-9 or AAV9), AAV type 10 (AAV-10 or AAV 10 or AAVrh10), avian AAV, bovine AAV, canine AAV, caprine AAV, equine AAV, primate AAV, non-primate AAV, ovine AAV. “Primate AAV” refers to AAV that infects primates, “non-primate AAV” refers to AAV that infects non-primate mammals, “bovine AAV” refers to AAV that infects bovine mammals, etc.

[0044] The genomic sequences of various AAV serotypes, as well as the sequences of the native terminal repeats (TR), Rep proteins, and capsid subunits are known in the art. Such sequences can be found in the literature or public databases such as GenBank. See, for example, GenBank accession numbers NC_002077.1 (AAV1), AF063497.1 (AAV1), NC_001401.2 (AAV2), AF043303.1 (AAV2), J01901.1 (AAV2), U48704.1 (AAV3A), NC_001729.1 (AAV3A), AF028705.1 (AAV3B), NC.001829.1 (AAV4), U89790.1 (AAV4), NC_006152.1 (AA5), AF085716.1 (AAV-5), AF028704.1 (AAV6), NC 006260.1 (AAV7), AF513851.1 (AAV8), AY530579.1 (AAV9), AAT46337 (AAV10), and AAO88208 (AAVrh10); these disclosures are incorporated herein by reference for teaching AAV polynucleotide and amino acid sequences.

[0045] As used herein, the term "recombinant adeno-associated virus capsid protein" or "rAAV capsid protein" refers to an AAV capsid protein that contains a 7 - amino acid peptide insertion in the GH loop of the VP1 - VP3 capsid proteins as compared to the wild - type VP1 - VP3 capsid proteins.

[0046] As used herein, the terms "recombinant adeno - associated virus particle", "rAAV virus particle", "rAAV vector", or "rAAV particle" refer to a virus particle that contains a recombinant / variant capsid protein.

[0047] If the AAV vector / virus particle contains a heterologous polynucleotide sequence, the heterologous polynucleotide sequence refers to a polynucleotide sequence that is not the wild - type AAV genome, e.g., a transgene delivered to a target cell, an RNAi agent delivered to a target cell, or a CRISPR agent, etc. Generally, the heterologous polynucleotide sequence is flanked by at least one, and usually two, AAV inverted terminal repeats (ITR).

[0048] As used herein, the term "heterologous" refers to an entity that is derived from a gene that is typically different from the remainder of the entity with which it is being compared. For example, a polynucleotide introduced into a plasmid or vector from a different species by genetic engineering techniques is a heterologous polynucleotide. Thus, for example, an rAAV containing a heterologous nucleic acid sequence encoding a foreign gene product is an rAAV that contains a polynucleotide that is not typically included in a naturally occurring wild-type AAV, and the encoded heterologous gene product is a gene product that is not typically encoded by a naturally occurring wild-type AAV.

[0049] As used herein, the term "packaging" refers to a series of intracellular events that result in the assembly and encapsulation of AAV particles. The AAV "rep" and "cap" genes refer to polynucleotide sequences that encode adenovirus-associated virus replication and capsidation proteins. AAV rep and cap are referred to herein as AAV "packaging genes".

[0050] As used herein, the term "polynucleotide" refers to a polymeric form of nucleotides of any length, including deoxyribonucleotides or ribonucleotides, or their analogs. Polynucleotides may include modified nucleotides, such as methylated nucleotides and nucleotide analogs, and may be interrupted by non-nucleotide components. If present, modifications to the nucleotide structure may be made before or after polymerization of the molecule. The term polynucleotide, as used herein, refers interchangeably to double-stranded and single-stranded molecules. Unless otherwise specified or required, any embodiment herein that includes a polynucleotide includes both the double-stranded form and each of the two complementary single-stranded forms that are known or predicted to constitute the double-stranded form.

[0051] As used herein, the term "gene" refers to a polynucleotide that performs a certain function in a cell. For example, a gene may contain an open reading frame capable of encoding a gene product. An example of a gene product is a protein, which is transcribed and translated from the gene. Another example of a gene product is RNA, such as a functional RNA product, such as an aptamer, interfering RNA, ribosomal RNA (rRNA), transfer RNA (tRNA), non-coding RNA (ncRNA), nuclease guide RNA, etc., which is transcribed but not translated.

[0052] Regarding "CRISPR / Cas9 agents", the term "CRISPR" encompasses Clustered Regularly Interspaced Short Palindromic Repeats / CRISPR-associated (Cas) systems, which provide bacteria and archaea with adaptive immunity against viruses and plasmids by using CRISPR RNAs (crRNAs) to direct the silencing of invading nucleic acids. The Cas9 protein (or its functional equivalents and / or variants, i.e., Cas9-like proteins) naturally contains DNA endonuclease activity, the activity of which depends on the binding of the protein to two naturally occurring or synthetic RNA molecules, called crRNA and tracrRNA, also known as guide RNAs. In some cases, these two molecules are covalently linked to form a single molecule (also known as single-guide RNA ("sgRNA")). Thus, Cas9 or a Cas9-like protein binds to DNA-targeting RNA (the term includes both the bimolecular guide RNA structure and the single-molecule guide RNA structure), activating the Cas9 or Cas9-like protein and guiding the protein to the target nucleic acid sequence. If the Cas9 or Cas9-like protein retains its native enzymatic function, it will cleave the target DNA to produce a double-strand break, resulting in genomic alterations (i.e., editing: deletions, insertions (when a donor polynucleotide is present), substitutions, etc.), thereby altering gene expression.

[0053] As used herein, the term "CRISPR agent" includes any agent (or nucleic acid encoding such an agent) that includes a natural or synthetic sequence and can be used in a Cas9-based system (e.g., Cas9 or a Cas9-like protein; any component of DNA-targeting RNA, e.g., crRNA-like RNA, tracrRNA-like RNA, single-guide RNA, etc.; donor polynucleotides; and the like).

[0054] As used herein, the terms "treatment", "treating", etc. refer to obtaining the desired pharmacological and / or physiological effect. For the purposes of preventing a disease or its symptoms completely or in part, the effect can be prophylactic, and / or for the purposes of curing a disease and / or an adverse effect attributable to the disease in part or completely, the effect can be therapeutic.

[0055] The terms "individual", "host", "subject", and "patient" are used interchangeably herein and refer to a mammal, including but not limited to humans; non-human primates, including apes; mammalian livestock (such as horses); farm mammals (such as sheep, goats, etc.); mammalian pets (dogs, cats, etc.); and rodents (such as mice, rats, etc.).

[0056] As used herein, the term "microglia" refers to cells of mesodermal / mesenchymal origin that migrate into the central nervous system and become resident macrophages in the unique brain microenvironment. Microglia are highly dynamic cells that interact with neurons and non-neuronal cells. Microglia patrol the brain parenchyma by continuous processes extension and retraction and are capable of transitioning from a branched morphology to an amoeboid morphology, a feature consistent with cell activation. Microglia express a wide range of receptors and thus respond to a variety of stimuli from neurotransmitters to cytokines and plasma proteins. They play a crucial role as regulators of synaptic function and phagocytosis of newborn neurons in the healthy brain and are important in synaptic plasticity and adult neurogenesis. In disease, they play a crucial role in neurological and neuroinflammatory disorders. Their interaction with T cells is a major component in the development of brain autoimmunity, while their pathogenic interaction with neurons through induction of ROS and iNOS plays a crucial role in neurological diseases. Emerging genetic tools and animal models have revealed the origin of microglia, their connection to peripheral monocytes, and their contribution to disease pathogenesis. Since microglia may play both beneficial and pathogenic functions in the central nervous system, understanding their role in disease-specific contexts is necessary for identifying novel microglia-targeted therapies for central nervous system diseases.

[0057] As used herein, the term "directed evolution" refers to a capsid engineering method that mimics natural evolution through repeated processes of genetic diversification and selection in vitro and / or in vivo, thereby accumulating beneficial mutations and progressively improving the function of biomolecules. Directed evolution typically involves an in vivo method called "biopanning" for selecting AAV variants from a library that have more efficient infectivity for a cell or tissue type of interest.

[0058] Regarding cell modification, the terms "gene modification" or "transformation" or "transfection" or "transduction" by exogenous DNA (e.g., by recombinant virus) refer to the introduction of such DNA into a cell. The presence of exogenous DNA results in a permanent or transient genetic change. The transforming DNA may or may not be integrated (covalently linked) into the genome of the cell.

[0059] Without being bound by theory, the present disclosure is in part based on the surprising discovery of several new AAV capsids that mediate efficient gene delivery to microglia through a screening process.

[0060] Recombinant adeno-associated virus (rAAV) vector

[0061] Adeno-associated virus (AAV) is a family of parvoviruses with a 4.7 kb single-stranded DNA genome within a non-enveloped capsid. The naturally occurring AAV viral genome has two inverted terminal repeats (ITRs) - the origin of viral replication and packaging signals - flanked by two major open reading frames (ORFs): rep (encoding proteins that function in viral replication, transcriptional regulation, site-specific integration, and virion assembly) and cap. The cap ORF encodes three structural proteins that assemble to form a 60-mer viral capsid. Many naturally occurring AAV variants and serotypes have been isolated, but none are associated with human disease.

[0062] Recombinant AAV (rAAV) can be used as a gene delivery vector, with a marker or a therapeutic gene of interest inserted between the ITRs, replacing rep and cap. These vectors have been shown to transduce both dividing and non-dividing cells in vitro and in vivo and can stably express the transgene for several years in post-mitotic tissues.

[0063] Recombinant AAV (rAAV) has shown promising results in an increasing number of clinical trials. However, barriers to gene delivery may limit the utility of AAV, such as anti-capsid immune responses, low transduction in certain tissues, inability to target delivery to specific cell types, and relatively low carrying capacity. In many cases, there is not enough mechanistic knowledge to effectively enable rational design to improve the capabilities of AAV. As an alternative, directed evolution has emerged as a strategy to create new AAV variants to meet specific biomedical needs. Directed evolution strategies utilize genetic diversification and selection processes to accumulate beneficial mutations, thereby gradually improving the function of biomolecules. In this process, the wild-type AAV cap gene is diversified by several methods to create large genetic libraries, which are packaged into viral particle libraries, and then selective pressure is applied to isolate new variants that can overcome the gene delivery barriers. Importantly, the mechanistic basis of the gene delivery problem does not need to be known for directed evolution of function, thus accelerating the development of enhanced vectors.

[0064] The three AAV capsid proteins, namely VP1, VP2, and VP3, are produced from the cap ORF in an overlapping manner through alternative mRNA splicing of the transcript and the use of alternative translation initiation codons. All three capsid proteins share a common stop codon. Although only VP1 is shown in the examples and figures, it should be understood that each of VP1, VP2, and VP3 contains the amino acid sequence of the 11 consecutive amino acids X1X2X3X4X5X6X7X8X9X 10 Q.

[0065] In some embodiments, compared to the wild-type VP1-VP3 capsid proteins, the 11 consecutive amino acids X1X2X3X4X5X6X7X8X9X provided by the present invention 10 The amino acid sequence of Q is inserted into the GH loop of the VP1, VP2, and / or VP3 capsid proteins.

[0066] For the VP1 capsid protein, the above amino acid sequence is inserted between the 588th amino acid and the 589th amino acid of the wild-type VP1 of AAV9, or at the corresponding position of the capsid protein of another AAV serotype different from AAV9. In some embodiments, the other AAV serotype may include AAV2, AAV1, AAV3A, AAV3B, AAV4, AAV5, AAV6, AAV7, AAV8, AAV10, etc. In some embodiments, the peptide insertion may be located between the 587th amino acid and the 588th amino acid of AAV2. In some embodiments, the peptide insertion may be located between the 591st amino acid and the 592nd amino acid of AAV1. In some embodiments, the peptide insertion may be located between the 588th amino acid and the 589th amino acid of AAV3A. In some embodiments, the peptide insertion may be located between the 588th amino acid and the 589th amino acid of AAV3B. In some embodiments, the peptide insertion may be located between the 584th amino acid and the 585th amino acid of AAV4. In some embodiments, the peptide insertion may be located between the 575th amino acid and the 576th amino acid of AAV5. In some embodiments, the peptide insertion may be located between the 591st amino acid and the 592nd amino acid of AAV6. In some embodiments, the peptide insertion may be located between the 589th amino acid and the 590th amino acid of AAV7. In some embodiments, the peptide insertion may be located between the 591st amino acid and the 592nd amino acid of AAV8. In some embodiments, the peptide insertion may be located between the 588th amino acid and the 589th amino acid of AAV10.

[0067] In some embodiments, the variant VP1 capsid protein, from the 587th amino acid to the 597th amino acid of AAV9, or at the corresponding position of the capsid protein of another AAV serotype different from AAV9, includes the 11 consecutive amino acids X1X2X3X4X5X6X7X8X9X provided by the present invention 10 The amino acid sequence of Q.

[0068] In some embodiments, the wild-type VP1 capsid protein of AAV9 includes the amino acid sequence shown in SEQ ID NO:6.

[0069] In some embodiments, the variant VP1 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:7. In some embodiments, the variant VP1 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:8. In some embodiments, the variant VP1 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:9. In some embodiments, the variant VP1 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:10. In some embodiments, the variant VP1 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:11. In some embodiments, the variant VP1 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:53. In some embodiments, the variant VP1 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:56. In some embodiments, the variant VP1 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:59. In some embodiments, the variant VP1 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:62. In some embodiments, the variant VP1 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:65. In some embodiments, the variant VP1 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:68.

[0070] In addition, according to specific embodiments, compared with the wild-type VP1 capsid protein or a variant VP1 capsid protein comprising the amino acid sequence shown in SEQ ID NO.:7, variant VP1 capsid proteins comprising the amino acid sequences shown in SEQ ID NO.:10, 11, 59, 62, 65 or 68 have significantly higher transduction ability for microglia in vivo. In addition, according to specific embodiments, compared with the wild-type VP1 capsid protein or a variant VP1 capsid protein comprising the amino acid sequence shown in SEQ ID NO.:8, variant VP1 capsid proteins comprising the amino acid sequences shown in SEQ ID NO.:9, 53 or 56 have significantly higher transduction ability for microglia in vivo.

[0071] For the VP2 capsid protein, the above 11 consecutive amino acids X1X2X3X4X5X6X7X8X9X 10The amino acid sequence of Q is inserted between the 451st and 452nd amino acids of the wild-type VP2 of AAV9, or at the corresponding position of the VP2 capsid protein of another AAV serotype different from AAV9. In some embodiments, the variant VP2 capsid protein, from the 450th amino acid to the 460th amino acid of AAV9, or the corresponding position of the VP2 capsid protein of another AAV serotype different from AAV9, comprises 11 consecutive amino acids X1X2X3X4X5X6X7X8X9X provided by the present invention. 10 The amino acid sequence of Q. Related definitions are the same as those described above for X1 to X 10

[0072] In some embodiments, the wild-type VP2 capsid protein of AAV9 comprises the amino acid sequence shown in SEQ ID NO:41.

[0073] In some embodiments, the variant VP2 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:43. In some embodiments, the variant VP2 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:45. In some embodiments, the variant VP2 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:47. In some embodiments, the variant VP2 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:49. In some embodiments, the variant VP2 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:51. In some embodiments, the variant VP2 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:54. In some embodiments, the variant VP2 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:57. In some embodiments, the variant VP2 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:60. In some embodiments, the variant VP2 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:63. In some embodiments, the variant VP2 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:66. In some embodiments, the variant VP2 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:69.

[0074] ​In addition, according to specific embodiments, compared with the wild-type VP2 capsid protein or a variant VP2 capsid protein comprising the amino acid sequence shown in SEQ ID NO.: 43, a variant VP2 capsid protein comprising the amino acid sequence shown in SEQ ID NO.: 49, 51, 60, 63, 66 or 69 has significantly higher transduction ability to microglia in vivo. In addition, according to specific embodiments, compared with the wild-type VP2 capsid protein or a variant VP2 capsid protein comprising the amino acid sequence shown in SEQ ID NO.: 45, a variant VP2 capsid protein comprising the amino acid sequence shown in SEQ ID NO.: 47, 54 or 57 has significantly higher transduction ability to microglia in vivo.

[0075] For the VP3 capsid protein, the above 11 consecutive amino acids X1X2X3X4X5X6X7X8X9X 10 The amino acid sequence of Q is inserted between the 386th amino acid and the 387th amino acid of the wild-type VP3 of AAV9, or at the corresponding position of the VP3 capsid protein of another AAV serotype different from AAV9. In some embodiments, the variant VP3 capsid protein, from the 385th amino acid to the 395th amino acid of AAV9, or at the corresponding position of the VP3 capsid protein of another AAV serotype different from AAV9, comprises the 11 consecutive amino acids X1X2X3X4X5X6X7X8X9X 10 The amino acid sequence of Q.

[0076] In some embodiments, the wild-type VP3 capsid protein of AAV9 comprises the amino acid sequence shown in SEQ ID NO: 42.

[0077] In some embodiments, the variant VP3 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:44. In some embodiments, the variant VP3 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:46. In some embodiments, the variant VP3 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:48. In some embodiments, the variant VP3 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:50. In some embodiments, the variant VP3 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:52. In some embodiments, the variant VP3 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:55. In some embodiments, the variant VP3 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:58. In some embodiments, the variant VP3 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:61. In some embodiments, the variant VP3 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:64. In some embodiments, the variant VP3 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:67. In some embodiments, the variant VP3 capsid protein of rAAV comprises the amino acid sequence shown in SEQ ID NO:70.

[0078] In addition, according to specific embodiments, compared with the wild-type VP3 capsid protein or a variant VP3 capsid protein comprising the amino acid sequence shown in SEQ ID NO.:44, variant VP3 capsid proteins comprising the amino acid sequences shown in SEQ ID NO.:50, 52, 61, 64, 67 or 70 have significantly higher transduction ability for microglia in vivo. In addition, according to specific embodiments, compared with the wild-type VP3 capsid protein or a variant VP3 capsid protein comprising the amino acid sequence shown in SEQ ID NO.:46, variant VP3 capsid proteins comprising the amino acid sequences shown in SEQ ID NO.:48, 55 or 58 have significantly higher transduction ability for microglia in vivo.

[0079] In some embodiments, compared with parental AAV virions that do not contain such a 7-amino acid peptide insertion, rAAV vectors / virions containing the above rAAV capsid proteins have significantly higher microglial transduction efficiency both in vitro and in vivo. In some specific embodiments, the microglia may be cultured microglia or microglia in the brain or spinal cord.

[0080] In addition, according to some embodiments, compared to wild-type VP1-VP3 capsid proteins or variant VP1-VP3 capsid proteins comprising the amino acid sequence shown in SEQ ID NO.:1, variant VP1-VP3 capsid proteins comprising the amino acid sequences shown in SEQ ID NO.:2, 33, 71, 72, 73 or 74 have significantly higher transduction ability for microglia in vivo.

[0081] In addition, according to some embodiments, compared to wild-type VP1-VP3 capsid proteins or variant VP1-VP3 capsid proteins comprising the amino acid sequence shown in SEQ ID NO.:4, variant VP1-VP3 capsid proteins comprising the amino acid sequences shown in SEQ ID NO.:5, 75 or 76 have significantly higher transduction ability for microglia in vivo.

[0082] Generally, the variants disclosed herein are generated by using one AAV library and / or multiple libraries. Such one or more AAV libraries are generated by mutating the cap gene encoding the structural proteins (e.g., VP1-VP3) of the AAV capsid, and are generated by a series of directed evolution techniques known and readily available to those skilled in the art of viral genome engineering, such as Cre-recombinase-based AAV targeted evolution (CREATE).

[0083] The CREATE process used herein is capable of developing AAV capsids that more efficiently transduce defined populations of cells expressing Cre in vivo or in vitro. To implement the CREATE strategy, Cre recombinase targeting sequences are added to the AAV variant genomes in the library. Without limitation, the CREATE process includes delivering the AAV library to cells or transgenic mice, including target cells that selectively express Cre recombinase. For those rAAVs that successfully transduce the target cells, their genomes will be modified by Cre recombinase. For those rAAVs that transduce non-target cells, their genomes will not be modified. In this way, specific primers can selectively recover the genomes of AAV variants capable of transducing target cells.

[0084] Once one or more AAV libraries are generated, the viruses are packaged such that each AAV particle consists of mutant VP1, VP2, and / or VP3 capsid proteins. The variants of the library are then subjected to in vitro and / or in vivo selection pressure techniques, which are well-known and readily available to those skilled in the art of AAV. For example, but not limited to, AAV variants can be selected by using: i) affinity columns, where elution of different fractions yields variants with altered binding properties; ii) primary cells - isolated from tissue samples or immortalized cell lines, mimicking the behavior of human cells, to yield AAV variants with higher efficiency and / or tissue specificity; iii) animal models - mimicking the clinical gene therapy setting - to yield AAV variants that successfully infect the target tissue; iv) human xenograft models, to yield AAV variants that infect transplanted human cells; and / or combinations of the selected techniques.

[0085] Once the viruses are selected, they can be recovered by known techniques, such as, but not limited to, adenovirus-mediated replication, PCR amplification, next-generation sequencing, and cloning, etc. The viral clones are then enriched by repeated selection techniques, and AAV DNA is isolated to recover the selected variant cap gene of interest. These selected variants can be subjected to further modification or mutation and thus can serve as a new starting point for further selection steps to iteratively increase the fitness of the AAV virus. However, in some cases, the successfully generated capsids have no additional mutations.

[0086] The AAV variants disclosed herein are at least partially generated by using in vitro or in vivo directed evolution methods, such as the techniques described above, involving screening in cultured primary mouse microglial cells or in vivo microglial cells after injection into the mouse striatum and / or midbrain. Thus, the AAV variant capsids disclosed herein include a 7-amino acid peptide inserted into the GH loop of VP1, VP2, and / or VP3, which confers more efficient transduction than the corresponding parental AAV capsid protein or a control group. As used herein, "corresponding parental AAV capsid protein" refers to an AAV capsid protein having the same wild-type or variant AAV serotype as the subject variant AAV capsid protein, but not containing the peptide insertion of the subject variant AAV capsid protein.

[0087] In some embodiments, the variant AAV capsid protein comprises a heterologous peptide consisting of 7 amino acids inserted covalently into the G-H loop or loop IV of the AAV capsid protein. By the "G-H loop" or loop IV of the AAV capsid protein is meant the solvent-accessible portion known in the art as the GH loop or loop IV of the AAV capsid protein.

[0088] In certain embodiments, the insertion site is a single insertion site between two adjacent amino acids located between amino acid 570 and amino acid 614 of VP1, amino acid 451 and amino acid 452 of VP2, and / or amino acid 386 and amino acid 387 of VP3 of any wild-type AAV serotype or AAV variant.

[0089] For example, the insertion site is between two adjacent amino acids at positions 570-610, 580-600, 570-575, 575-580, 580-585, 585-590, 590-600, or 600-614 of any AAV serotype or variant VP1. In a preferred embodiment, the insertion site is between the 580th and 581st amino acids, between the 581st and 582nd amino acids, between the 583rd and 584th amino acids, between the 584th and 585th amino acids, between the 585th and 586th amino acids, between the 586th and 587th amino acids, between the 587th and 588th amino acids, between the 588th and 589th amino acids, or between the 589th and 590th amino acids. The insertion site can be between the 575th and 576th amino acids, between the 576th and 577th amino acids, between the 577th and 578th amino acids, between the 578th and 579th amino acids, or between the 579th and 580th amino acids. The insertion site can be between the 590th and 591st amino acids, between the 591st and 592nd amino acids, between the 592nd and 593rd amino acids, between the 593rd and 594th amino acids, between the 594th and 595th amino acids, between the 595th and 596th amino acids, between the 596th and 597th amino acids, between the 597th and 598th amino acids, between the 598th and 599th amino acids, or between the 599th and 600th amino acids. In a preferred embodiment, the insertion site is between the 587th and 588th amino acids of AAV2, between the 591st and 592nd amino acids of AAV1, between the 588th and 589th amino acids of AAV3A, between the 588th and 589th amino acids of AAV3B, between the 584th and 585th amino acids of AAV4, between the 575th and 576th amino acids of AAV5, between the 591st and 592nd amino acids of AAV6, between the 589th and 590th amino acids of AAV7, between the 591st and 592nd amino acids of AAV8, between the 588th and 589th amino acids of AAV9, or between the 588th and 589th amino acids of AAV10.Those skilled in the art will know that, based on the comparison of the amino acid sequences of various AAV serotype capsid proteins, the insertion site of "amino acids corresponding to AAV9" will be located in the capsid protein of any given AAV serotype.

[0090] The peptides disclosed herein have not been previously described and / or inserted into the AAV capsid. Without wishing to be bound by theory, the presence of any disclosed peptide insertion may increase the transduction of AAV into microglia in vitro or in vivo.

[0091] In certain embodiments, for the amino acid sequence X1X2X3X4X5X6X7X8X9X 10 Q, X1 is selected from Ala or Leu; X2 is selected from Gln, Met, Thr, Val or Pro; X3 is selected from Trp, Thr, Glu, Pro, Leu, Ala or Gln; X4 is selected from Pro, Thr, Met, Ser, Arg or Ala; X5 is selected from Pro, Ser, Val, Asp or Phe; X6 is selected from Lys or Pro; X7 is selected from Thr or Arg; X8 is selected from Thr, Glu or Pro; X9 is selected from Ser, Pro or Ala; X 10 is selected from Ala or Asp.

[0092] In certain embodiments, the variant VP1, VP2, and / or VP3 capsid proteins comprise the amino acid sequences AQWPPKTTSAQ (SEQ ID NO.:1), LMTPPKTTSAQ (SEQ ID NO.:2), ATEPPKTTSAQ (SEQ ID NO.:3), AQQRPPREPAQ (SEQ ID NO.:4), AQQRPPRPADQ (SEQ ID NO.:5), AQPTSKTTSAQ (SEQ ID NO.:71), AQLMVKTTSAQ (SEQ ID NO.:72), AQWTDKTTSAQ (SEQ ID NO.:73), AVLSPKTTSAQ (SEQ ID NO.:74), AQTAFPREPAQ (SEQ ID NO.:75), or APARPPREPAQ (SEQ ID NO.:76).

[0093] In some embodiments, the rAAV viral particles can be produced by co-transfecting 293 / 293T cells with a plasmid expressing the variant capsid protein of the present disclosure, an adenovirus helper plasmid, and optionally a transgenic plasmid expressing a heterologous polynucleotide sequence.

[0094] In some embodiments, the rAAV viral particles comprise a heterologous polynucleotide sequence encoding a gene product. In some embodiments, the gene product is an interfering RNA. In some embodiments, the gene product is a long or short non-coding RNA. In some embodiments, the gene product is an antisense RNA. In some embodiments, the gene product is a guide RNA. In some embodiments, the gene product is an aptamer. In some embodiments, the gene product is a polypeptide. In some embodiments, the gene product is a secreted antibody. In some embodiments, the gene product is a single-chain antibody. In some embodiments, the gene product is a VHH domain. In some embodiments, the gene product is a soluble receptor. In some embodiments, the gene product is an affibody. In some embodiments, the gene product is a chaperone. In some embodiments, the gene product is a site-specific nuclease that provides site-specific knockdown of gene function.

[0095] In some embodiments, compared to the parental AAV viral particles and other wild-type AAV serotypes, the rAAV viral particles of the present disclosure have higher transduction efficiency and stronger heterologous polynucleotide sequence expression.

[0096] In some embodiments, the rAAV viral particles of the present disclosure are transduced into cultured microglial cells without inducing pro-inflammatory pathways in the microglial cells.

[0097] In certain embodiments, compared to the parental AAV9 viral particles and other wild-type AAV serotypes, the rAAV9 viral particles comprising variant VP1, VP2, and / or VP3 capsid proteins have higher transduction efficiency and stronger heterologous polynucleotide sequence expression in cultured microglial cells, wherein the variant VP1, VP2, and / or VP3 capsid proteins comprise the amino acid sequence of AQWPPKTTSAQ (SEQ ID NO.:1) or AQQRPPREPAQ (SEQ ID NO.:4).

[0098] In certain embodiments, compared to rAAV9 viral particles comprising AQQRPPREPAQ (SEQ ID NO.: 1), parental AAV9 viral particles, and other wild-type AAV serotypes, rAAV9 viral particles comprising variant VP1, VP2, and / or VP3 capsid proteins further enhance the transduction efficiency of microglia in vivo and stronger heterologous polynucleotide sequence expression, wherein the variant VP1, VP2, and / or VP3 capsid proteins comprise the amino acid sequences of LMTPPKTTSAQ (SEQ ID NO.: 2), ATEPPKTTSAQ (SEQ ID NO.: 3), AQPTSKTTSAQ (SEQ ID NO.: 71), AQLMVKTTSAQ (SEQ ID NO.: 72), AQWTDKTTSAQ (SEQ ID NO.: 73), or AVLSPKTTSAQ (SEQ ID NO.: 74).

[0099] In certain embodiments, compared to rAAV9 viral particles comprising AQQRPPREPAQ (SEQ ID NO.: 4), parental AAV9 viral particles, and other wild-type AAV serotypes, rAAV9 viral particles comprising variant VP1, VP2, and / or VP3 capsid proteins further enhance the transduction efficiency of microglia in vivo and stronger heterologous polynucleotide sequence expression, wherein the variant VP1, VP2, and / or VP3 capsid proteins comprise the amino acid sequences of AQQRPPRPADQ (SEQ ID NO.: 5), AQTAFPREPAQ (SEQ ID NO.: 75), or APARPPREPAQ (SEQ ID NO.: 76).

[0100] In some embodiments, the expression of rAAV-mediated transgenes in microglia can be further increased by pharmacological means. In certain embodiments, topoisomerase and proteasome inhibitors are used to further enhance the expression level of heterologous nucleotide sequences transduced using the rAAV viral particles of the present disclosure. In a preferred embodiment, a topoisomerase inhibitor, such as doxorubicin, is used to increase the expression level of heterologous nucleotide sequences transduced using the rAAV viral particles of the present invention.

[0101] In some embodiments, transduction mediated by the rAAV viral particles of the present disclosure does not induce microglial activation either in vitro or in vivo.

[0102] In some embodiments, the rAAV viral particles of the present disclosure carry a heterologous nucleotide sequence encoding a CRISPR agent for genome editing in vitro and in vivo, such as gene knockout. In some embodiments, the rAAV viral particles of the present disclosure carry a heterologous nucleotide sequence encoding a guide RNA.

[0103] The following examples are intended to provide a complete disclosure and description to those of ordinary skill in the art to guide how to prepare and use the variant AAV capsids disclosed herein, and are not intended to limit the scope of the invention disclosed herein. In addition, the following examples do not represent that the following experiments are all or the only experiments.

[0104] Methods

[0105] Mice. In accordance with the Regulations for the Administration of Laboratory Animals in China, the health care and use of animals were approved by the Animal Management and Use Committee of the National Institute of Biological Sciences, Beijing. Cx3cr1 CreER mice (021160, B6.129P2(Cg)-Cx3cr1tm2.1(cre / ERT2)Litt / WganJ), Cx3cr1GFP mice (005582, B6.129P2(Cg)-Cx3cr1tm1Litt / J), and Rosa26-LSL-Cas9 mice (024857, B6;129-Gt(ROSA)26Sortm1(CAG-cas9*, -EGFP)Fezh / J) were obtained from Jackson Laboratory. Adult mice of either sex were used for in vivo virus injection. Day 1 postnatal (P1) and adult C57BL / 6N wild-type mice were from Beijing Vital River Laboratory Animal Technology Co., Ltd. Mice were maintained on a 12 / 12-hour light cycle (lights on at 8:00 am) and provided with food and water ad libitum.

[0106] Plasmids. Plasmids for capsid screening were constructed as follows: The pAAV-CMV-mScarlet-ΔCap-DIO-SV40pA plasmid contains the mScarlet expression cassette, the cis Cap cassette, and the DIO cassette. The mScarlet expression cassette consists of a CMV promoter, the mScarlet coding sequence, and the SV40pA sequence. The cis Cap cassette includes the AAV5 p41 promoter sequence, the AAV2 rep splicing sequence, and the AAV9 cap sequence. The AAV9 cap sequence was modified to introduce XbaI and AgeI sites for subsequent library generation. The DIO expression cassette contains the SV40pA sequence. The pCRII-9Cap-xE plasmid and the AAV2 / 9REP-AAP helper plasmid were constructed according to the original report. To facilitate library construction, the 449th amino acid lysine of wild-type VP1, the 312th amino acid of wild-type VP2, and / or the 247th amino acid of wild-type VP3 were mutated to arginine.

[0107] The pAAV-SFFV backbone was constructed by replacing the Ef1a promoter of pAAV-DIO-hChR2(H134R)-mCherry (Addgene 20297) with the SFFV promoter of the pAAV-SFFV plasmid (Addgene 46911). The mScarlet coding sequence was subcloned into the DIO expression cassette to prepare the pAAV-SFFV-DIO-mScarlet plasmid, or subcloned behind the SFFV promoter to prepare the pAAV-SFFV-mScarlet plasmid. To generate the pAAV-SFFV-DIO-jGCaMP8s plasmid, the jGCaMP8s (Addgene 162380) coding sequence was synthesized and subcloned into the DIO expression cassette. The hSyn-Cre-2A-GFP-KASH expression cassette on the original plasmid (Addgene 60231) was replaced with the DIO-mScarlet expression cassette to construct the pAAV-U6-sgRNA-SFFV-DIO-mScarlet plasmid. sgRNAs targeting Tmem119 (5′-GGGACCCCGTACCTTCAGCG) and Cd68 (5′-atcctatacccaattcaggg) were selected from the mouse Brie CRISPR KO sgRNA library (Addgene 73632), then synthesized and cloned into the pAAV-U6-sgRNA-SFFV-DIO-mScarlet plasmid. The SaCas9 sgRNA (5’-CGGCTCCCAGTTTAGCATCACT) targeting P2ry12 was designed using the web tool Benchling (https: / / benchling.com / crispr), then synthesized and cloned into the original pX601 plasmid 45 according to the SaCas9 user manual.

[0108] To minimize potential spontaneous recombination during bacterial growth, the Stbl3 cell line was used to amplify the AAV vector containing the DIO expression cassette.

[0109] AAV packaging. The AAV vector was packaged according to a protocol commonly used in the art. Briefly, the AAV vector and AAV helper plasmids were co-transfected into HEK293T cells. Cells were harvested 96 hours after transfection, and viral particles were released from the cells by freeze-thaw cycles and sonication. The virus was purified by cesium chloride density gradient ultracentrifugation and dialyzed into phosphate-buffered saline (PBS) buffer. The virus titer was detected by qPCR.

[0110] Isolation and culture of mouse microglia. Primary mouse microglia were derived from P1 C57BL / 6 wild-type mice. The pups were placed on ice for 1 - 2 min until unresponsive, then soaked in 75% alcohol and the heads were carefully removed. The brains were collected with clean sterile scissors and placed in a 10-cm culture dish containing 10 mL of ice-cold dissociation medium [DMEM / F12 (11330032, Gibco), supplemented with 100 U / mL penicillin and 100 μg / mL streptomycin (P / S, 15140 - 122, Gibco)]. The meninges were removed with Dumont forceps No. 5 under a dissecting microscope. The brains were mechanically dissociated in the dissociation medium. The dissociated cells were filtered through a 40-μm cell strainer and centrifuged at 1000 rpm for 10 min at room temperature. The microspheres were resuspended in medium [DMEM / F12, supplemented with 10% fetal bovine serum (FBS, 0099 - 141, Gibco), 5 ng / mL granulocyte-macrophage colony-stimulating factor (GM-CSF, PRP100489, Abbkine) and 1% P / S] and plated at a density of five brains per T-75 plastic culture flask (Falcon) coated with poly-L-lysine (P8920, Sigma-Aldrich). The medium was changed 24 h after isolation. Thereafter, 50% of the medium was changed every 3 days. After two weeks, the flasks were shaken on an orbital shaker at 180 rpm for 2 h at 37 °C to harvest the microglia. The cultured microglia were maintained in a humidified incubator at 37 °C with 5% CO2.

[0111] In vitro AAV transduction. To test the candidate capsids identified from in vitro screening, microglia were placed in 96-well cell culture plates (6005550, PerkinElmer). The microglia were transduced with rAAV packaged with the candidate capsids at a multiplicity of infection (MOI) of 10,000. Doxorubicin (0.1 μg / mL; D1515, Sigma-Aldrich) was added to the medium before rAAV transduction. After 2 days, the medium was changed to TIC medium [DMEM / F12 supplemented with 1% P / S, 2 mM L-glutamine (25030-081, Gibco), 5 mg / mL N-acetylcysteine (A9165, Sigma-Aldrich), 5 mg / mL insulin (I0516, Sigma-Aldrich), 100 mg / mL transferrin (T1147, Sigma-Aldrich), 100 ng / mL sodium selenite (S5261, Sigma-Aldrich), 2 ng / mL recombinant mouse TGF-β2 (50153-M08H, Sino Biological), 100 ng / mL recombinant mouse IL-34 (50055-M08H, Sino Biological), 1.5 mg / mL cholesterol (wool, 700000P, Merck)]. Five days after rAAV transduction, fluorescence imaging and analysis were performed using an Opera Phenix high-content screening system (PerkinElmer).

[0112] RNA sequencing of cultured microglia. For M1 microglia polarization, mouse primary microglia were exposed to 200 ng / mL lipopolysaccharide (LPS, L4130, Sigma-Aldrich) in TIC medium for 24 hours. For M2 microglia polarization, mouse primary microglia were exposed to 20 ng / mL recombinant mouse interleukin-4 (214-14, PeproTech) for 24 hours. Total RNA was extracted from untreated control, LPS-treated, interleukin-4-treated, and rAAV-transduced microglia using Trizol (15596018, Thermo Fisher Scientific) and subjected to single-end 75-bp high-throughput sequencing on an Illumina platform.

[0113] In vitro screening. The detailed sequences of the primers used in this study are listed in Table 1. AAV-cMG.WPP and AAV-cMG.QRP were identified by screening AAV variants that could efficiently transduce cultured mouse microglia. First, an AAV capsid library was constructed by inserting random heptamers into the reading frames of each capsid protein VP 1-3 of the AAV9 capsid using the CREATE protocol. Briefly, the pCRII-9Cap-xE plasmid was used as a template, and the library fragment was generated by PCR with XF and 7xMNN primers. The pAAV-CMV-mScarlet-ΔCap-DIO-SV40pA plasmid was linearized with XbaI and AgeI. The library fragment was assembled into the linearized pAAV-CMV-mScarlet-ΔCap-DIO-SV40pA plasmid by Gibson assembly. The resulting library was packaged into rAAV by co-transfecting HEK293T cells with the AAV capsid library, AAV2 / 9REP-AAP helper plasmid, and AAV helper plasmid. Approximately 10 library rAAVs were used to transduce cultured mouse microglia for 24 hours. After 48 hours of transduction, the rAAV genome of successfully transduced cultured microglia was recovered using Trizol. First, the cap sequence was amplified from the recovered AAV genome by PCR with 9CapF and SV40pA-R primers. The purified PCR product was used as a template for a second PCR reaction with XF and 588i-R primers. Then the recovered cap sequence was assembled back into the pAAV-CMV-mScarlet-ΔCap-DIO-SV40pA plasmid and screened again in cultured mouse microglia. The highly enriched candidates after two rounds of screening were determined by next-generation sequencing (NGS) and tested individually. The enrichment score of the variant was calculated as follows:

[0114] Enrichment score = Log 10 ((Normalized read count in the second round) / (Normalized read count in the first round)).

[0115] To further identify the AAV-cMG.WPP and AAV-cMG.QRP variants that mediate more efficient microglial transduction in vitro, AAV-cMG.WPP and AAV-cMG.QRP capsid mutant libraries were constructed, in which the inserted heptamers and four flanking amino acids in the AAV-cMG.WPP or AAV-cMG.QRP capsid were randomized ( Figure 1 C and 2D). The AAV-cMG.WPP and AAV-cMG.QRP capsid mutant libraries were screened in cultured mouse microglia as described above.

[0116]

[0117]

[0118] In vivo screening. To identify AAV-cMG.WPP and AAV-cMG.QRP variants that efficiently transduce mouse microglia in vivo, AAV-cMG.WPP and AAV-cMG.QRP capsid mutant libraries were constructed in which the inserted heptamers and four flanking amino acids in the AAV-cMG.WPP capsid were randomized ( Figure 1 C and 2D). Briefly, using the pCRII-9Cap-xE plasmid as a template, 10 separate PCR reactions were performed with XF and WPP-mut-R1-10 primers to generate library fragments. Equal amounts of the 10 PCR products were mixed and assembled into the pAAV-CMV-mScarlet-ΔCap-DIO-SV40pA plasmid using Gibson assembly. As described above, the resulting library was packaged into rAAV. The rAAV of the AAV-cMG.WPP and AAV-cMG.QRP capsid mutant libraries were injected into the striatum (800 nL) and midbrain (500 nL) of 3 Cx3cr1 CreER mice. Tamoxifen (intraperitoneal injection, 10 mg / kg) was injected continuously for 5 days after virus injection. The mice were sacrificed 10 days after virus injection. The brains were dissected, and the rAAV genomes of the cells successfully transduced in vivo were recovered using Trizol. The cap sequence in the Cre-recombined genome was selectively amplified using 9CapF and CDF primers. Highly enriched candidates were identified by NGS and tested individually. The enrichment score of the variant was calculated as follows:

[0119] Enrichment score = Log 10 ((Normalized read count of in vivo screening sample) / (Normalized read count of AAV library)).

[0120] Screening NGS sample preparation. The library rAAV was treated with proteinase K overnight at 37 °C. The AAV genome was obtained by phenol-chloroform extraction and ethanol precipitation. 100 ng of the AAV genome was used as a PCR template. The cap fragment containing the inserted heptamer was amplified using 588i-F and 588i-R primers. The PCR product was purified and used as a template for a second PCR, which was amplified using 1527 and 1532 primers. Finally, PCR was performed to add unique indices for subsequent NGS using standard primers. Before submitting to NGS, the size of the indexed PCR product was selected using 2% agarose gel.

[0121] In in vivo selection, the cap sequence in the Cre-recombined AAV genome was selectively amplified using 9CapF and CDF primers. The cap fragment containing the inserted heptamer was then amplified and indexed as described above.

[0122] General surgery and virus injection. Mice were anesthetized with pentobarbital (intraperitoneal injection, 80 mg / kg) before surgery and then placed in a mouse stereotaxic apparatus. Then, the virus was injected into the orbitofrontal cortex (OFC), striatum, midbrain, hippocampus, or thalamus of the mice.

[0123] Injections were performed using a microinjection pump (Nanoliter 2010 syringe, WPI) and a Micro4 controller (WPI). The virus was delivered to the target area at a rate of 46 nL / min. Immediately after virus injection, doxorubicin (150 ng / g) was injected retroorbitally. All subsequent experiments were performed at least 2 weeks after virus injection, except for experiments involving in vivo CRISPR / Cas9 KO, which were performed at least 3 weeks after virus injection. For Cx3cr1 CreER mice, tamoxifen was injected intraperitoneally for 3 consecutive days after virus injection (in vivo screening was performed for 5 consecutive days as described above).

[0124] Single-cell isolation was used for microglial single-cell RNA sequencing (scRNA-seq). To minimize in vitro activation, a cold-mechanical dissociation protocol was employed. All procedures were performed on ice with cold buffers or in a refrigerated centrifuge. Mice were deeply anesthetized and perfused. The brain was quickly removed and immersed in Dounce buffer (HBSS containing HEPES + DNase + RNase inhibitor), and cut into smaller pieces. The tissue solution was quickly transferred to a 15-ml Dounce homogenizer and gently homogenized about 10 times with a loose pestle. The remaining tissue pieces were pelleted, and the supernatant containing the cell suspension was collected into a new tube. Fresh Dounce buffer was added to the pelleted tissue, and another round of homogenization was repeated. The collected cell suspension was centrifuged, resuspended, and passed through pre-wetted cell filters of 70 μm and 30 μm in sequence to remove debris. The cells were centrifuged again and resuspended in 37% isotonic Percoll (SIP). Microglia were enriched with a Percoll gradient of HBSS / 30% / 37% (cells) / 70%, centrifuged at 200 g for 20 min with minimum acceleration and no braking. The interphase cells between 30% and 37% were carefully collected, washed, and resuspended in Dulbecco's phosphate-buffered saline (Dulbecco's PBS) containing 0.04% BSA. Then, the cells were manually picked for Smart-seq or run through the 10’ Genomics Chromium Single Cell 3'v3 protocol, followed by sequencing on an Illumina platform.

[0125] Construction and sequencing of scRNA-seq libraries based on Smart-seq2. Single microglial cells in bovine serum albumin (BSA) buffer [HBSS buffer with 0.04% BSA (0332, Amresco)] were aspirated under a stereofluorescence microscope using a pipette. As previously described, an improved STRT-seq protocol was used to construct single-cell RNA-seq libraries for library construction. Briefly, cells were lysed, and mRNA was released into the lysis buffer containing barcoded reverse transcription primers. The mRNA was reverse transcribed into cDNA. After pre-amplification of the cDNA, single-cell transcriptomes with different barcode labels were pooled together. Another 4 PCR cycles were performed using biotinylated primers to enrich the 3' ends of the cDNA. The amplicons were randomly sheared into fragments of approximately 300 bp by sonication (Covaris) and purified using Dynabeads MyOne streptavidin C1 (65002, Invitrogen). The purified fragments were sequenced with 150 bp paired-end reads on the Illumina platform.

[0126] Sequencing data processing and quality control (QC). The reference dataset generated by the 10’Genomics platform was aligned with the pre-built GRCm38 reference genome by "cellranger count" and combined by the "cellranger aggr" command (Cell Ranger v3.1.0). The resulting feature barcode matrix file was loaded with Seurat41 (v4.0.2) software and filtered according to the following QC criteria: cells with mitochondrial RNA greater than 10% and cells expressing more than 300 features were removed, and features expressed in fewer than 10 cells were removed. After QC, the Seurat workflow was run with default parameters, and the CellSelector function of UMAP embedding was used to manually subset the clusters showing persistent high expression of Hexb58 to retain true microglial cells from the dataset.

[0127] For the Smart-seq2 dataset, the protocol listed in the uni-tools documentation (https: / / umi-tools.readthedocs.io / en / latest / Single_cell_tutorial.html) was adopted, where fastq files were demultiplexed using uni-tools59, aligned with STAR60 (v2.5.4b), and quantified using featureCounts61 (Subread v1.6.3). A slight modification was made to the "umi_tools whitelist" process, where 100 cells were searched to obtain an extended whitelist, but the whitelist was "cleaned" by only retaining the cell barcodes actually used in the library preparation step, which were named "ground truth barcode". In the "umi_tools extract" step, only those reads containing the "ground truth barcode" were retained, while reads with adapter contamination or major sequencing errors were removed. Library size, number of features, mitochondrial RNA percentage, and ribosomal RNA percentage in the feature cell matrix were calculated for quality control. The scater62 (v1.18.6) quickPerCellQC function was used to identify and remove cell outliers based on these four metrics. Transduction of rAAV in sequenced microglia was determined by detecting mScarlet reporter transcripts.

[0128] Unsupervised analysis of microglial state dimensionality reduction and identification. The Seurat workflow was performed with default parameters on the 10’ reference dataset and the Smart-seq2 dataset. Standard PCA was calculated for both datasets, and the dimensionality of the datasets was selected as 40 (dims = 1:40). To determine the microglial state, microglia from control mice were first labeled as "steady state", and microglia from LPS-treated mice were labeled as "responsive state". Then, the FindTransferAnchors function was used to find transfer anchors between the query (Smart-seq2) and reference (10’) datasets in the PCA space, and the MapQuery function was used to project the query data onto the reference dataset. The reference dataset and the query dataset were further merged, and UMAP dimensionality reduction was re-run on the merged dataset to obtain a "de novo visualization" in the UMAP space.

[0129] In vivo two-photon imaging. At 8 weeks of age, rAAV (1 μL) packaged with AAV-MG1.1 or AAV-MG1.2 capsids was injected into Cx3cr1 CreEROr in layers II-IV of the primary somatosensory cortex (S1) of Cx3cr1GFP mice. Two weeks later, a circular skull with a diameter of 3 mm centered on the virus injection site was carefully removed. A custom-designed steel head bar including the imaging chamber was placed on the craniotomy and fixed to the exposed skull with cyanoacrylate glue and dental cement. The mice were allowed to recover from the surgery for one week. Two-photon imaging was performed at 100-150 μm below the dura mater using a FluoView FVMPE-RS microscope (Olympus, 25×, 1.05 NA water immersion lens). The laser was tuned to 920 nm, and the jGCaMP8s and GRABATP imaging were maintained below 50 mW, and the GFP imaging was maintained below 40 mW. Microglia were imaged using z-stacks, with each z-stack consisting of 16 images spaced 2 μm apart (total depth 30 μm). Lateral offsets were corrected using the StackReg plugin. Data quantification was performed using a custom Matlab script and GraphPad Prism software.

[0130] To monitor the changes in jGCaMP8s and GRAB ATP1.0 fluorescence signals after LPS stimulation, imaging was performed at 1.5 fps on a 636×636 μm field of view (512×512 pixel resolution, 1.24 μm / pixel). First, 10 min of time-lapse imaging (50 z-stacks) was performed to record the baseline fluorescence. Then, LPS (10 mg / kg) was injected intraperitoneally. Another 10 min of imaging (6 times for jGCaMP8s imaging and ATP1.0 7 times for GRAB imaging) was performed immediately after LPS injection, and imaging was performed every 50 min. A maximum projection was created for each stack. Microglia were manually selected. During the 10 min of imaging, the fluorescence intensity of the microglial cell body was the average of 25 z-stacks.

[0131] To monitor the changes in GRAB ATP1.0 fluorescence signals after laser ablation, imaging was performed at 1.5 fps on a 636×636 μm field of view (512×512 pixel resolution). A circular area with a diameter of 15 μm at the center of the imaging field was focused, and a 920 nm laser pulse of approximately 500 mW (30 ms / pulse) was applied to induce acute tissue damage. 40 min of time-lapse imaging (200 z-stacks) was performed immediately after laser ablation. A maximum projection was created for each stack. The fluorescence intensity was calculated by manually selecting the microglial cell body.

[0132] For imaging the response of GFP-expressing microglia to acute tissue injury, the field of view of 318×318 μm (512×512 pixel resolution, 0.62 μm / pixel) was imaged at 1.5 fps. As described above, acute tissue injury was induced. Time-lapse imaging was performed immediately after laser ablation, at 40 min (control mice; 200 z-stack) or 80 min (P2ry12 knockout mice; 400 z-stack). The normalized microglial response was calculated as described above.

[0133] Immunohistochemistry. Mice were anesthetized with an overdose of pentobarbital, perfused intracardially with PBS, and then perfused with paraformaldehyde (PFA, 4% wt / vol in PBS). At room temperature, the brains were dissected and fixed in 4% PFA for at least 4 h. Then the samples were dehydrated in 30% sucrose solution. Brain sections (30 μm) were prepared on a Cryostat microtome (Leica CM1950). The sections were permeabilized with 0.3% Triton X-100 in PBS (PBST) and blocked with PBST containing 2% BSA for 1 h at room temperature. The sections were incubated with primary antibodies (anti-Iba1, 1:500, 019-19741, Wako; anti-TMEM119, 1:1000, ab209064, Abcam; anti-CD68, 1:500, ab53444, Abcam; anti-P2RY12, 1:100, 848002, BioLegend) overnight at 4 °C. The samples were washed three times in PBST and then incubated with fluorescent secondary antibodies (goat anti-rabbit-AF647, 111-605-144, Jackson ImmunoResearch; goat anti-rabbit-AF488, 111-545-003, Jackson ImmunoResearch; goat anti-rat-Cy5, 112-175-143, Jackson ImmunoResearch) for 2 h at room temperature.

[0134] For cultured microglia, the cells were first washed with cold PBS and then fixed with 4% PFA for 10 min at room temperature. After washing again in PBS, the cells were permeabilized in PBST and blocked with PBST containing 2% BSA for 20 min at room temperature. Then the cells were incubated with an antibody (anti-Iba1, 1:500, 019-19741, Wako) for 2 h at room temperature. The cells were washed 3 times in PBST and then incubated with a fluorescent secondary antibody (goat anti-rabbit-AF488, 111-545-003, Jackson ImmunoResearch) for 1 h at room temperature.

[0135] Image acquisition and analysis of fixed tissues. Confocal microscopy observations were performed on a Zeiss LSM880 confocal scanning microscope. For slide scanning imaging, wide-field fluorescence imaging was performed using an Olympus VS120 virtual microscope slide scanning system with a 10’ objective lens. Cell counting was performed using Imaris. To evaluate the effect of doxorubicin administration on mScarlet expression, the average fluorescence intensity of labeled cells in the striatum of 9 brain slices centered on the viral injection site was measured using Imaris. To quantify the knockout efficiency of Tmem119 sgRNA, the percentage of signal-positive pixels in a 1.5 mm × 1.5 mm area covering the dorsal striatum was calculated using the "Threshold" function in ImageJ. To quantify the knockout efficiency of P2ry12 sgRNA, the percentage of signal-positive pixels on a 1 mm × 1 mm area covering the S1 cortex was calculated using the "Threshold" function in ImageJ. In the Cd68 knockout experiment, the number of Cd68-positive cells was quantified using Imaris in a 1.5 mm × 1.5 mm area covering the dorsal striatum.

[0136] Example

[0137] Example 1. In vitro screening of capsid libraries

[0138] The wild-type AAV9 capsid was used as the starting point for generating the capsid library, where each AAV9 capsid variant contained a random 7-amino acid insertion between amino acids 588 and 589 of the AAV9 VP1 protein ( Figure 1 A). The library was packaged into rAAV and continuously screened in cultured mouse microglia for two rounds ( Figure 1 A and 2A). Cultured mouse microglia were transduced with the capsid library rAAV, and the capsid variants that successfully mediated transduction were recovered. Then, the recovered capsid variants were packaged into rAAV and screened again in cultured mouse microglia. Through next-generation sequencing, the highly enriched capsid variants after two rounds of screening were identified ( Figure 1 B and 2B).

[0139] Two capsid variants, one containing the "WPPKTTS" heptamer insertion (hereinafter referred to as AAV-cMG.WPP; Figure 1 B), and one containing the "QRPPREP" heptamer insertion (hereinafter referred to as AAV-cMG.QRP; Figure 2B), transduction of cultured microglia was significantly higher compared to other candidates tested. The VP1 protein of AAV-cMG.WPP has the amino acid sequence as shown in SEQ ID NO.:7, and the VP1 protein of AAV-cMG.QRP has the amino acid sequence as shown in SEQ ID NO.:8. Then, the single-stranded mScarlet reporter vector was packaged into rAAV with the candidate capsid variants and transduced into cultured mouse microglia. The transduction ability of these capsid variants was compared with the parental AAV9 capsid and three AAV capsids that have been reported to transduce cultured mouse microglia [AAV5, AAV8, and AAV6 with the triple mutation Y731F / Y705F / T492V (AAV6TM)28].

[0140] In the second round of screening, AAV-cMG.WPP was enriched more than 170-fold and accounted for 12.91% of the total recovered variants ( Figure 1 B). The transduction rate of AAV-cMG.WPP (~75%) was significantly increased compared to the AAV5 (~12%), AAV6TM (~3%), AAV8 (~34%), or AAV9 (~10%) capsids ( Figure 3 A-C). AAV-cMG.WPP also significantly enhanced the expression of mScarlet compared to the AAV5, AAV6TM, AAV8, or AAV9 capsids ( Figure 3 A and C).

[0141] In the second round of screening, AAV-cMG.QRP was enriched approximately 400-fold and accounted for 5.05% of the total recovered variants ( Figure 2 B). AAV-cMG.QRP had significantly higher transduction rates and stronger mScarlet expression compared to the AAV5, AAV6TM, AAV8, or AAV9 capsids ( Figure 2 C, E, G).

[0142] Example 2. Examination of the effect of AAV-cMG.WPP on microglial phenotype

[0143] External stimuli can induce microglia to enter a reactive or alternative activated state. To examine whether AAV-cMG.WPP-mediated transduction can trigger microglial phenotype changes, principal component analysis was performed on transcriptome data obtained from four different samples: untransduced control (steady state), LPS treatment (reactive state), interleukin-4 treatment (alternative activated state), and AAV-cMG.WPP-transduced cultured mouse microglia ( Figure 3 D). AAV-cMG.WPP-transduced microglia aggregated towards untransduced microglia in the control group and away from microglia treated with LPS or interleukin-4 ( Figure 3D). Further differential gene expression analysis also indicated that AAV-cMG.WPP transduction did not induce pro-inflammatory pathways in cultured microglia Figure 3 E). Thus, these results demonstrated the utility of AAV-MG.WPP in mediating safe transgene delivery in cultured microglia.

[0144] Example 3. Further screening of the AAV-cMG.WPP semi-random mutant capsid library for in vivo transduction

[0145] An additional capsid library was generated by semi-random mutagenesis of the heptamer inserted in AAV-cMG.WPP and its adjacent four amino acids Figure 1 C). This new library was packaged into rAAV and subjected to in vivo screening by injecting library rAAV into the brains of Cx3cr1 CreER mice. The CREATE strategy was used to selectively recover capsid variants from Cre-recombined AAV genomes (i.e., rAAV genomes that had successfully transduced microglia in vivo). After two rounds of screening, two highly enriched capsid variants were identified Figure 1 D), both of which contained mutations at amino acid positions 587-589 of AAV-cMG.WPP. The first variant included the amino acid sequence "LMT" at positions 587-589, accounting for 13.8% of the total recovered variants Figure 1 D). The second variant included the amino acid sequence "ATE" at positions 587-589, accounting for 5.7% of the total recovered variants Figure 1 D). These two AAV-cMG.WPP capsid variants were named AAV-MG1.1 and AAV-MG1.2, respectively. The VP1 protein of AAV-MG1.1 has the amino acid sequence shown in SEQ ID NO.:10. The VP1 protein of AAV-MG1.2 has the amino acid sequence shown in SEQ ID NO.:11.

[0146] Four other AAV-cMG.WPP variants were also identified to be able to transduce microglia in vivo and induce strong and widespread mScarlet expression in the striatum of Cx3cr1 CreER mice Figure 12 C-F). The first variant, AAV-MG.PTS, included the amino acid sequence "PTS" at positions 589-591 of AAV-cMG.WPP Figure 12 C). The second variant, AAV-MG.LMV, included the amino acid sequence "LMV" at positions 589-591 of AAV-cMG.WPP Figure 12D). The third variant, AAV-MG.WTD, includes the amino acid sequence "WTD" at positions 589-591 of AAV-cMG.WPP ( Figure 12 E). The fourth variant, AAV-MG.VLS, includes the amino acid sequence "VLS" at positions 588-590 of AAV-cMG.WPP ( Figure 12 F). The VP1 protein of AAV-MG.PTS has the amino acid sequence shown in SEQ ID NO.:59. The VP1 protein of AAV-MG.LMV has the amino acid sequence shown in SEQ ID NO.:62. The VP1 protein of AAV-MG.WTD has the amino acid sequence shown in SEQ ID NO.:65. The VP1 protein of AAV-MG.VLS has the amino acid sequence shown in SEQ ID NO.:68.

[0147] Example 4. Testing the ability of AAV-MG1.1 and AAV-MG1.2 to transduce microglia in vivo

[0148] Single-stranded Cre-dependent mScarlet reporter vectors were packaged into rAAV (AAV-MG1.1 / 1.2-SFFV-DIO-mScarlet) using AAV-MG1.1 and AAV-MG1.2. Then, these rAAV were injected into the brains of Cx3cr1 CreER mice, and the corresponding AAV-cMG.WPP rAAV were evaluated. Both AAV-MG drove strong and widespread mScarlet expression in all tested brain regions (orbitofrontal cortex, striatum, midbrain, hippocampus, and thalamus); Figure 1 E, 4A, 4B, and 5A).

[0149] Further studies found that using the immunoreactivity of ionized calcium-binding adapter molecule 1 (Iba1) as a marker for microglia, the expression of mScarlet was restricted to microglia, and almost all microglia in the core region of the injection site were labeled ( Figure 1 F, 5C, and 5D). The in vivo microglial transduction efficiency of both AAV-MG was also much higher than that of the parental AAV9 capsid and much higher than that of the AAV5 and AAV6TM capsids. In Cx3cr1 CreER mice, AAV9, AAV5, or AAV6TM rAAV only achieved sparse and faint labeling of microglia ( Figure 5 E).

[0150] Example 5. Testing the effect of topoisomerase inhibitors on the in vivo microglial transduction efficiency of AAV-MG

[0151] Previous studies have confirmed that the use of small molecule drugs to inhibit topoisomerase and proteasome can promote rAAV transduction in vitro and in vivo. Therefore, it was examined whether this method could further improve the efficiency of microglial transduction of AAV-MG in vivo. The FDA-approved topoisomerase inhibitor doxorubicin has been shown to increase the expression level of rAAV in neurons in vivo. In Cx3cr1 CreER After stereotaxic injection of AAV-MG1.1 / 1.2-SFFV-DIO-mScarlet into the brains of Cx3cr1 mice, doxorubicin was immediately administered via retro-orbital injection. For both AAV-MGs, doxorubicin administration significantly increased the expression level of mScarlet in microglia, but did not increase the number of mScarlet-labeled microglia ( Figure 1 E, 5A and 5B). These results demonstrated the possibility of using pharmacological methods to further enhance the expression of AAV-mediated transgenes in microglia.

[0152] Example 6. Examine the effect of AAV-MG-mediated transduction on microglial activation in vivo

[0153] After determining the effective transduction of microglia by AAV-MG in vivo, it was examined whether AAV-MG-mediated transduction would trigger phenotypic changes in microglia in the brain. Two weeks after injecting AAV-MG1.1 / 1.2-SFFV-DIO-mScarlet into the striatum of Cx3cr1 CreER mice, mScarlet + and mScarlet - microglia were selected and single-cell RNA sequencing (scRNA-seq) was performed using an improved Smart-seq2 protocol. Projection of microglia using Uniform Manifold Approximation and Projection (UMAP) showed that untransduced and transduced microglia formed homogeneous clusters ( Figure 6 A and 7A). In addition, there were no differences in the expression of homeostatic microglial marker genes or reactive microglial marker genes between untransduced and transduced microglia ( Figure 6 B). Using the label transfer strategy implemented in Seurat, microglia in the Smart-seq2 dataset were projected onto the reference 10× datasets of homeostatic and reactive microglia ( Figure 7 B-D), and the combined UMAP projection was recalculated to represent the microglial states from both datasets ( Figure 6 C). Most microglia in the Smart-seq2 dataset were located within the homeostatic microglial cluster ( Figure 6 C). These results supported that AAV-MG-mediated transduction did not induce microglial activation in vivo.

[0154] Example 7. Analysis of the Delivery Efficiency of AAV-MG into Microglia

[0155] In this example, it was tested whether AAV-MG could effectively deliver various genetic payloads into microglia. Two newly developed gene-encoded fluorescent sensors were expressed in microglia to examine the physiological responses of microglia to peripheral endotoxin challenge ( Figure 8 A). First, a single-stranded Cre-dependent AAV vector carrying the latest generation calcium indicator jGCaMP8s (AAV-MG1.2-SFFV-DIO-GCaMP8s) was packaged using AAV-MG1.2. The rAAV was injected into the primary somatosensory (S1) cortex of Cx3cr1 CreER mice. After allowing viral transgene expression for two weeks, 10 mg / kg LPS (i.p.) was injected to induce systemic inflammation. The Ca 2+ signal in microglial cell bodies was tracked by two-photon imaging ( Figure 8 A). The results showed that the Ca 2+ signal increased significantly 1 hour after LPS injection ( Figure 8 B and 8C). This LPS-induced intracellular Ca 2+ continued to rise and reached a peak 3 hours after injection, then began to decline ( Figure 8 B and 8C). The increase in the Ca 2+ signal was not caused by the intraperitoneal injection process itself, because intraperitoneal injection of saline did not result in significant changes in jGCaMP8s fluorescence ( Figure 8 B and 8C).

[0156] Furthermore, a single-stranded Cre-dependent AAV vector containing the newly developed ATP fluorescent sensor GRAB ATP1.0 was packaged using AAV-MG1.2 (AAV-MG1.2-SFFV-DIO-GRAB ATP1.0 ). The GRAB ATP1.0 sensor was expressed in microglia of the S1 cortex of Cx3cr1 CreER mice, and the GRAB fluorescence signal was monitored by in vivo two-photon imaging. After intraperitoneal injection of 10 mg / kg LPS, the changes in extracellular ATP of microglial cell bodies were detected ( Figure 8 A). Similar to the above GCaMP imaging experiment, the GRAB fluorescence signal increased significantly 1 hour after LPS injection ( Figure 8 D and 8E). The fluorescence signal tended to stabilize 2 hours after injection and remained at a high level 6 hours after injection ( Figure 8D and 8E). The results showed that the continuous increase in ATP signal in microglial cell bodies after LPS injection was different from the LPS-induced ATP blinks observed in astrocytes, indicating that cells have specific ATP sensing after peripheral endotoxin stimulation. After intraperitoneal injection of saline, no obvious signal changes were seen ( Figure 8 D and 8E), again indicating that the elevated GRAB fluorescence signal was not attributed to the intraperitoneal injection procedure.

[0157] The changes in extracellular ATP in microglial cell bodies during acute tissue injury were also examined. Transient laser ablation was used to induce local tissue injury, resulting in an immediate increase in GRAB fluorescence signal, which lasted for up to 40 min ( Figure 9 ).

[0158] Example 8. Analysis of the genome editing ability of AAV-MG in microglia

[0159] This example explored the utility of AAV-MG in microglial genome editing. Cre-dependent Rosa26 Cas9 reporter mice (Rosa26-LSL-Cas9) were crossed with Cx3cr1 CreER mice. Tamoxifen injection induced the expression of Streptococcus pyogenes Cas9 (SpCas9) in microglia. AAV-MG was used to package single-stranded AAV vectors expressing sgRNA. Two genes, Tmem119 and Cd68, selectively expressed in brain microglia were selected as editing targets (AAV-MG1.1-sgRNA-Tmem119 and AAV-MG1.2-sgRNA-Cd68). rAAV carrying sgRNA targeting LacZ was used as a control virus (AAV-MG1.1 / 1.2-sgRNA-LacZ). Four weeks after injecting rAAV into the striatum, its knockout effect was examined by immunostaining. For both target genes, the immunofluorescence signal of their encoded proteins was significantly reduced in the striatum of mice injected with knockout rAAV, but not in the brains of mice injected with control rAAV ( Figure 10 A, 10B, 11A and 11B). Effective knockout of the target genes was achieved in a large area of the dorsal striatum along the anteroposterior axis by single intrastriatal injection of AAV ( Figure 10 A, 10B, 11A and 11B). These results indicate that combining Cas9 transgenic mice with AAV-MG-mediated sgRNA delivery can perform in vivo genome editing of microglia in a region-specific manner.

[0160] Example 9. Demonstration of the ability of AAV-MG to directly edit the microglial genome in vivo

[0161] To further demonstrate direct in vivo microglial genome editing, miniaturized Cas9 (Staphylococcus aureus Cas9, SaCas9) and sgRNA from a single AAV vector (AAV-MG1.2-CMV-SaCas9-U6-sgRNA) were expressed in microglia using AAV-MG1.2. The sgRNA was designed to target the microglial homeostasis marker gene P2ry12, which is essential for extracellular polynucleotide-activated microglia. From the absence of immunostaining signals, it can be seen that in the S1 cortex of Cx3cr1 GFP transgenic mice, a single injection of this rAAV can largely knockout the P2ry12 gene, while Cx3cr1 GFP transgenic mice selectively express GFP in microglia ( Figure 11 C and 11D). It is known that microglial immune activation can downregulate the expression of P2ry12. To rule out the possibility that the observed P2ry12 knockout is due to rAAV-triggered microglial activation, a control rAAV (AAV-MG1.2-CMV-SaCas9) carrying only SaCas9 without sgRNA was used. The results showed that the expression of SaCas9 without sgRNA did not affect the expression of endogenous P2ry12 ( Figure 11 C and 11D), indicating that AAV-MG1.2-mediated transduction does not induce microglial activation and that SaCas9 gene knockout requires a specific sgRNA.

[0162] Example 10. Examination of the Physiological Consequences of AAV-MG1.2-Mediated P2ry12 Gene Knockout

[0163] To examine the physiological consequences of AAV-MG1.2-mediated P2ry12 gene knockout, AAV-MG1.2-CMV-SaCas9-U6-sgRNA-P2ry12 or AAV-MG1.2-CMV-SaCas9 vector was injected into the S1 cortex of Cx3cr1GFP transgenic mice. Four weeks later, the morphological response of microglia to tissue damage was tracked by in vivo two-photon imaging of the S1 cortex ( Figure 10 C). Focal laser ablation of the mouse brain injected with the control rAAV induced rapid extension and recruitment of microglial processes to the damaged site, reaching a peak 30 - 40 min after ablation ( Figure 10 D and 10E). In sharp contrast, the chemotactic response of microglia in the brains of mice injected with the P2ry12 knockout rAAV was significantly reduced: the microglial response only began approximately 30 min after local laser ablation and reached a peak approximately 80 min later ( Figure 10D and 10E). The AAV-based gene knockout method circumvents the potential confounding effects of P2ry12 deficiency during development, and thus these results further support the role of P2ry12 in guiding the directional branch extension of microglia to the cortical injury site.

[0164] Example 11. Further screening of the AAV-cMG.QRP semi-random mutant capsid library

[0165] To further improve the transduction efficiency of AAV-cMG.QRP. In cultured mouse microglia, another round of directed evolution of AAV-cMG.QRP was performed ( Figure 2 D). First, an additional capsid library was generated by semi-random mutagenesis of the inserted heptamer and four adjacent amino acids in AAV-cMG.QRP ( Figure 2 D). This new library was packaged into rAAV and screened in cultured mouse microglia. After one round of screening, a highly enriched capsid variant was identified that contains mutations at amino acids 594-596 of AAV-cMG.QRP ( Figure 2 D). This variant includes the amino acid sequence "PAD" at positions 594-596 and accounts for 0.79% of the total recovered variants ( Figure 2 D). This variant was named AAV-cMG. Compared with the AAV5, AAV6TM, AAV8, AAV9, or AAV-cMG.QRP capsids, AAV-cMG has significantly higher transduction efficiency and stronger mScarlet expression ( Figure 2 E, 2F, and 2G). Similarly, doxorubicin also significantly increased the mSaclet expression level of AAV-cMG in cultured mouse microglia ( Figure 2 H and 2I). The VP1 protein of AAV-cMG has the amino acid sequence shown in SEQ ID NO.:9.

[0166] In Cx3cr1 CreER The AAV-cMG.QRP mutant library was screened in the mouse brain. Two variants capable of transducing microglia in vivo were identified. The first variant, AAV-MG.TAF, includes the amino acid sequence "TAF" at positions 589-591 of AAV-cMG.QRP ( Figure 12 A). The second variant, AAV-MG.APA, includes the amino acid sequence "APA" at positions 587-589 of AAV-cMG.QRP ( Figure 12 B). The VP1 protein of AAV-MG.TAF has the amino acid sequence shown in SEQ ID NO.:53, and the VP1 protein of AAV-MG.APA has the amino acid sequence shown in SEQ ID NO.:56.

Claims

1. A recombinant adeno-associated virus (rAAV) capsid protein comprising an amino acid sequence of 11 consecutive amino acids X1X2X3X4X5X6X7X8X9X 10 Q, wherein X1 is selected from Ala or Leu; X2 is selected from Gln, Met, Thr, Val or Pro; X3 is selected from Trp, Thr, Glu, Pro, Leu, Ala or Gln; X4 is selected from Pro, Thr, Met, Ser, Arg or Ala; X5 is selected from Pro, Ser, Val, Asp or Phe; X6 is selected from Lys or Pro; X7 is selected from Thr or Arg; X8 is selected from Thr, Glu or Pro; X9 is selected from Ser, Pro or Ala; and X 10 is selected from Ala or Asp.

2. The rAAV capsid protein according to claim 1, wherein X6 is Lys; X7 is Thr; X8 is Thr; X9 is Ser; and / or X 10 is Ala.

3. The rAAV capsid protein according to claim 2, wherein X1 is selected from Ala or Leu; X2 is selected from Gln, Met, Thr or Val; X3 is selected from Trp, Thr, Glu, Pro or Leu; X4 is selected from Pro, Thr, Met or Ser; and / or X5 is selected from Ser, Val, Asp or Pro.

4. The rAAV capsid protein according to claim 2 or 3, wherein X1 is Ala, X2 is Gln, X3 is Trp, X4 is Pro, X5 is Pro; X1 is Leu, X2 is Met, X3 is Thr, X4 is Pro, X5 is Pro; X1 is Ala, X2 is Thr; X3 is Glu, X4 is Pro, X5 is Pro; X1 is Ala, X2 is Gln; X3 is Pro, X4 is Thr, X5 is Ser; X1 is Ala, X2 is Gln; X3 is Leu, X4 is Met, X5 is Val; X1 is Ala, X2 is Gln; X3 is Trp, X4 is Thr, X5 is Asp; or X1 is Ala, X2 is Val; X3 is Leu, X4 is Ser, X5 is Pro.

5. The rAAV capsid protein according to any one of claims 1-4, wherein the capsid protein comprises an amino acid sequence selected from the following: AQWPPKTTSAQ (SEQ ID NO.:1), LMTPPKTTSAQ (SEQ ID NO.:2), ATEPPKTTSAQ (SEQ ID NO.:3), AQPTSKTTSAQ (SEQ ID NO.:71), AQLMVKTTSAQ (SEQ ID NO.:72), AQWTDKTTSAQ (SEQ ID NO.:73), and AVLSPKTTSAQ (SEQ ID NO.:74).

6. The rAAV capsid protein according to claim 1, wherein X1 is Ala; X6 is Pro; and / or X7 is Arg.

7. The rAAV capsid protein according to claim 6, wherein X2 is selected from Gln or Pro; X3 is selected from Thr, Ala or Gln; X4 is selected from Arg or Ala; X5 is selected from Pro or Phe; X8 is selected from Glu or Pro; X9 is selected from Pro or Ala; and / or X 10 is selected from Ala or Asp.

8. The rAAV capsid protein according to claim 6 or 7, wherein X2 is Gln, X3 is Gln, X4 is Arg, X5 is Pro, X8 is Glu, X9 is Pro, X 10 is Ala; X2 is Gln, X3 is Gln, X4 is Arg, X5 is Pro, X8 is Pro, X9 is Ala, X 10 is Asp; X2 is Gln, X3 is Thr, X4 is Ala, X5 is Phe, X8 is Glu, X9 is Pro, X 10 is Ala; or X2 is Pro, X3 is Ala, X4 is Arg, X5 is Pro, X8 is Glu, X9 is Pro, X 10 is Ala.

9. The rAAV capsid protein according to any one of claims 6-8, wherein the capsid protein comprises an amino acid sequence selected from the following: AQQRPPREPAQ (SEQ ID NO.:4), AQQRPPRPADQ (SEQ ID NO.:5), AQTAFPREPAQ (SEQ ID NO.:75), and APARPPREPAQ (SEQ ID NO.:76).

10. The rAAV capsid protein according to any one of claims 1-9, wherein compared with the wild-type capsid protein, the 11 consecutive amino acids X1X2X3X4X5X6X7X8X9X 10 The amino acid sequence of Q is inserted into the GH-loop, preferably inserted between the 588th and 589th amino acids of wild-type VP1 of AAV9, between the 451st and 452nd amino acids of wild-type VP2, and / or between the 386th and 387th amino acids of wild-type VP3, or at the corresponding position of the capsid protein of another AAV serotype different from AAV9.

11. A polynucleotide sequence encoding the rAAV capsid protein according to any one of claims 1-10.

12. A recombinant adeno-associated virus (rAAV) vector comprising the rAAV capsid protein according to any one of claims 1-10.

13. The rAAV vector according to claim 12, further comprising a heterologous polynucleotide sequence, wherein the heterologous polynucleotide sequence preferably encodes a heterologous polypeptide, non-coding RNA, or CRISPR agent.

14. The rAAV vector according to claim 13, wherein the heterologous polynucleotide sequence encodes miRNA, siRNA, piRNA, lncRNA, or guide RNA.

15. A pharmaceutical composition comprising the rAAV vector according to any one of claims 12-14, and one or more pharmaceutically acceptable carriers.

16. A method for delivering the rAAV vector according to any one of claims 12-14 to a target cell, comprising contacting the target cell with the rAAV vector.

17. The method according to claim 16, wherein the target cell is a microglial cell in vitro or in vivo, preferably in the presence of an inhibitor of topoisomerase, proteasome, or non-homologous end joining, or a DNA damage inducer, and the inhibitor is more preferably selected from doxorubicin, bortezomib, etoposide, teniposide, vanillin, and bleomycin.

18. A host cell comprising the rAAV vector according to any one of claims 12-14.

19. A method for treating a neurological disease, comprising administering to a subject in need a therapeutically effective amount of the pharmaceutical composition according to claim 15, wherein the neurological disease is a disease associated with microglial cells.

20. The method according to claim 19, wherein the neurological disease includes Alzheimer's disease, Parkinson's disease, Huntington's disease, amyotrophic lateral sclerosis, leukoencephalopathy, glioma, and atypical teratoid / rhabdoid tumor.

Images