The therapeutic implications of all-in-one AAV-delivered epigenome-editing platform in neurodegenerative disorders

Abstract

Safely and efficiently controlling gene expression is a long-standing goal of biomedical research, and CRISPR/Cas system can be harnessed to create powerful tools for epigenetic editing. Adeno-associated-viruses (AAVs) represent the delivery vehicle of choice for therapeutic platform. However, their small packaging capacity isn't suitable for large constructs including most CRISPR/dCas9-effector vectors. Thus, AAV-based CRISPR/Cas systems have been delivered via two separate viral vectors. Here we develop a compact CRISPR/dCas9-based repressor system packaged in AAV as a single optimized vector. The system comprises the small Staphylococcus aureus (Sa)dCas9 and an engineered repressor molecule, a fusion of MeCP2's transcription repression domain (TRD) and KRAB. The dSaCas9-KRAB-MeCP2(TRD) vector platform repressed robustly and sustainably the expression of multiple genes-of-interest, in vitro and in vivo, including ApoE, the strongest genetic risk factor for late onset Alzheimer's disease (LOAD). Our platform broadens the CRISPR/dCas9 toolset available for transcriptional manipulation of gene expression in research and therapeutic settings.

Fig. 1. Improvement and optimization of the viral backbone.

a Sp1 and NF-kB binding sites were introduced upstream from EFS-NC promoter (pEFS-NC) (construct 1). The backbone integrated with 2xSp1 (two yellow circles) is outlined in construct 2. 4xSp1 (four yellow circles) is highlighted in construct 3. The backbone integrated with 2xNF-kB (two red circles) inserted upstream from pEFS-NC is outlined in construct 4. The backbone carrying 4xNF-kB (four red circles) is highlighted in construct 5. 2xSp1/2xNF-kB is depicted in construct 6. The complete EF1a promoter-driven plasmid is outlined in construct 7. The vector carried eGFP – nano-Luciferase reporter. b Physical titer of the modified AAV vectors. The lane order is the same as in the legend to (c). The experiment was done in triplicates. The statistical analysis was done using Prism GraphPad software. Ordinary one-way ANOVA was carried out to determine significant difference in the samples’ means ± SEM (P = 0.0002). ns P > 0.05, *P ≤ 0.05, **P ≤ 0.01. ***P ≤ 0.001, ****P ≤ 0.0001; multiple comparisons of means were determined using Tukey’s multiple comparisons test (c) The expression of the modified AAV vectors has been assessed in HEK293T cells. The Luciferase assay was used to determine Relative Light Units (RLU) signals. Lane 1: EFS-NC core promoter- harboring vector packaged into AAV2.9 particles. Lane 2: EFS-NC core promoter- harboring 2xSp1 vector packaged into AAV2.9 particles. Lane 3: EFS-NC core promoter- harboring 4xSp1vector packaged into AAV2.9 particles. Lane 4: EFS-NC core promoter- harboring 2xNF-kB vector packaged into AAV2.9 particles. Lane 5: EFS-NC core promoter- harboring 4xNF-kB vector packaged into AAV2.9 particles. Lane 6: EFS-NC core promoter- harboring 2xSp1 and 2xNF-kB vector packaged into AAV2.9 particles. Lane 7: EF1-alpha complete promoter vector packaged into AAV2.9 particles. The experiment was done in quadruplicates. The statistical analysis was done using Prism GraphPad software. Ordinary one-way ANOVA was carried out to determine significant difference in the samples’ means ± SEM (P < 0.0001). ns P > 0.05, *P ≤ 0.05, **P ≤ 0.01. ***P ≤ 0.001, ****P ≤ 0.0001; multiple comparisons of means were determined using Tukey’s multiple comparisons test.

Fig. 2. Schematic representation of the screened repressors paired with dSaCas9 and dCjCas9.

a Schematic view of the LV-reporter vector and the parental AAV vector carrying various repressor effectors. LV harbors a dual reporter system consisting of destabilized GFP (dGFP) and Nano-Luciferase (NLuc). Viral long terminal repeats – LTRs; SV40 polyadenylation signal- SV40pA are highlighted.pSV40 drives the expression of the puromycin marker. The vector was transduced at the MOI = 0.2 to enable the selection of the cells carried 1 copy/cell. Woodchuck Hepatitis Virus (WHP) Posttranscriptional Regulatory Element (WPRE). The inverted terminal repeats (ITRs) of AAV are highlighted. The EFS-NC promoter that drives the expression of the dCas9-effector (black and red boxes, respectively) is highlighted. Human U6 promoter driving expression of gRNA is highlighted. b Schematic representation of Heterochromatin Protein 1 alpha (HP1a) and Heterochromatin Protein 1 beta (HP1b) The Chromodomain (CD) and Chromoshadow Domain (CSD) are highlighted. The Hinge region separating the CD and CSD domains is highlighted. c Schematic representation of Methyl-Binding Proteins (MBDs). Methyl-Binding Protein 1 (MBD1); Methyl-Binding Protein 2 (MBD2); Methyl-Binding Protein 3 (MBD3); and Methyl-CpG Binding Protein 2 (MeCP2) are highlighted. Methyl- Binding Domain (MBD) highlighted here, is responsible for the protein-DNA binding Transcription –Repression Domain (TRD) highlighted here is responsible for protein-protein interactions directly involved in gene silencing (d) Schematic representation of DNA methyltransferases (DNMTs). De novo methyl-transferase A and B (DNMT3A and B, respectively) are highlighted here. The PWWP domain, named for a conserved Pro-Trp-Trp-Pro motif is highlighted. The Methyl-transferase catalytic domains (MTase) of DNMT3A and DNMT3B are highlighted. e Schematic representation of Nuclear inhibitor of Protein phosphatase 1 (NIPP1) protein. Embryonic Ectoderm Development domain (EED) binding to Polycomb Repressive Complex 2 (PRC2) is highlighted. The Krüppel associated box (KRAB) consists of the repressive boxA and boxB domains which is schematically represented. f Schematic representation of CMV promoter targeted by AAV-KRAB-MeCP2(TRD) system. CjCas9 –matching gRNA1 and gRNA2 are highlighted. SaCas9 – matching gRNA1 and gRNA2 are highlighted. PAM of CjCas9 and SaCas9 are highlighted.

Fig. 3. Titers and expression of the engineered AAV vectors.

The titers of the a dCjCas9 and b dSaCas9-vectors. Lane 1: negative. Lane 2: HP1a-no-hinge/no-gRNA. Lane 3: HP1a-no-hinge/gRNA1. Lane 4: HP1a-no-hinge/gRNA2. Lane 5: HP1a-hinge/no-gRNA. Lane 6: HP1a-hinge/gRNA1. Lane 7: HP1a-hinge/gRNA2. Lane 8: HP1b/no-gRNA. Lane 9: HP1b-no-hinge-gRNA1. Lane 10: HP1b-no-hingle-gRNA2. Lane 11: HP1b-hinge/no-gRNA. Lane 12: HP1b-hinge/gRNA1. Lane 13: HP1b-hinge/gRNA2. Lane 14: MBD1/no-gRNA. Lane 15: MBD1/gRNA1. Lane 16: MBD1/gRNA2. Lane 17: MBD2/no-gRNA. Lane 17: MBD2/gRNA1. Lane 18: MBD2/gRNA2. Lane 19: MBD3/no-gRNA. Lane 20: MBD3/gRNA1. Lane 21: MBD3/gRNA2. Lane 22: NIPP1/no-gRNA. Lane 23: NIPP1/gRNA1. Lane 24: NIPP1/gRNA2. Lane 25: KRAB/no-gRNA. Lane 26: KRAB/gRNA1. Lane 27: KRAB/gRNA2. Lane 28: MeCP2/no-gRNA. Lane 29: MeCP2/gRNA1. Lane 30: MeCP2/gRNA2. Lane 31: KRAB-MeCP2/no-gRNA. Lane 32: KRAB-MeCP2/gRNA1. Lane 33: KRAB-MeCP2/gRNA2. Lane 34: DNMT3A/no-gRNA. Lane 35: DNMT3A/gRNA1. Lane 36: DNMT3A/gRNA2. Lane 37: DNMT3B/no-gRNA. Lane 38: DNMT3B/gRNA1. Lane 39: DNMT3B/gRNA2. b the lane order is the same as for dCjCas9 in (a). The statistical analysis was done using Prism GraphPad. Ordinary one-way ANOVA was carried out to determine significant difference in the samples’ means ± SEM. ns P > 0.05, *P ≤ 0.05, **P ≤ 0.01. ***P ≤ 0.001, ****P ≤ 0.0001; multiple comparisons of means were determined using Tukey’s test. Expression of the vectors: dCjCas9 (c) and dSaCas9 (d). The lane’s order is the same as for (a, b). Relative light units (RLU) were recorded. The statistical analysis, as above (c P < 0.0001, d P < 0.0001).

Fig. 4. Expression of AAV-dSaCas9-KRAB-MeCP2(TRD) vectors in the reporter-HEK293T and HEPG2 cells.

a The samples were harvested 1-,2-and 3-weeks-post-transduction, and the luciferase assay was performed. Weeks 1–3; Lanes 1,4,7: Naïve cells. Lanes 2,5,8: AAV-dSaCas9-/no-gRNA. Lanes 3,6,9: AAV-dSaCas9-KRAB-MeCP2(TRD)/gRNA1. The experiment was done in quadruplicates. The statistical analysis, as above. b dGFP-expression. The samples were harvested 1-,2-,3- and 4 weeks-post-transduction, and the Western Blot was performed using anti-GFP-Ab. Weeks 1–4: Lanes 1,7,13,19 Untransduced cells; Lanes 2,8,14,20 AAV-dSaCas9-/no-gRNA; Lanes 3,9,15,21 AAV-no-Sp1/NF-kB; dSaCas9-KRAB-MeCP2/gRNA1; Lanes 4,10,16,22 AAV-no-Sp1/NF-kB-dSaCas9-KRAB-MeCP2/gRNA1; Lanes 5,11,17,23 AAV-dSaCas9- Sp1/NF-kB/no-gRNA. Lanes 6,12,18,24 AAV-dSaCas9- Sp1/NF-kB-KRAB-MeCP2/gRNA1. One-sample T-tests were carried out across normalized data, determining significant differences in means compared to the theoretical mean of 1. The values of Lanes 1,7,13,19 were normalized to equal the theoretical mean of 1. ns P > 0.05, *P ≤ 0.05, **P ≤ 0.01. ***P ≤ 0.001, ****P ≤ 0.0001. The densitometry was measured using ImageJ software. c The samples were harvested 1-,2-,3- and 4 weeks-post-transduction, and the Western Blot was performed using anti-Pcsk9-Ab. Weeks 1–4; Lane 1,7,13,19. Untransduced HEPG2 cells. Lane 2,8,14,20. No-KRAB-MeCP2/no-gRNA. Lane 3,9,15,21. gRNA1 targeting Psck9 vector. Lane 4,10,16,22. gRNA2 targeting Pcsk9 vector. Lane 5,11,17,23. gRNA3 targeting Psck9 vector. The statistical analysis was done as above. The densitometry was measured using ImageJ.

Fig. 5. In vivo validation of the all-in-one AAV/ dSaCas9- KRAB-MeCP2(TRD) repressor platform in mouse hippocampus.

The all-in-one AAV/ dSaCas9- KRAB-MeCP2(TRD) repressor platform and the control vector AAV/dSaCas9 with no repressor were administered by stereotaxic injection into the mouse dorsal hippocampus (DH) and validated using a GFP reporter gene (a–c). a, b LV-GFP reporter vector was co-injected with the AAV/gRNA1-dSaCas9- KRAB-MeCP2(TRD) into the left DH and with the control AAV/dSaCas9 into the right DH. The AAV/gRNA1-dSaCas9- KRAB-MeCP2(TRD) vector repressed the expression of the GFP reporter gene. a Representative images of brain coronal slices at 2x magnification 14 days and 42 days post-injection, showing GFP expression and DAPI staining in the DH. b Signals were quantified using ImageJ. Box plot displays the ratios of the left DH relative to the right DH in both age groups of 16 weeks (14d post-injection n = 9, p = 0.003; 42d post-injection n = 11, p = 0.028) and 32 weeks (n = 10, p = 0.026) mice. Each open circle represents the quantified signal left/right for a mouse. c The AAV/gRNA1-dSaCas9- KRAB-MeCP2(TRD) vector repressed the expression of the GFP mRNA, box plot displays mean relative expression of GFP mRNA at 14 days (n = 5, p = 0.027) or 42 days (n = 6, p = 0.00046) post-injection. Each open circle represents the relative expression (log2) for a mouse. Values represent mean ± SEM. *p < 0.05 **p  <  0.01 ***p < 0.001; Two-tailed paired t-test. Source data are provided as a Source Data file.

Fig. 6. In vivo validation of the all-in-one AAV/ dSaCas9- KRAB-MeCP2(TRD) repressor platform using ApoE in mouse hippocampus.

The all-in-one AAV/ dSaCas9- KRAB-MeCP2(TRD) repressor platform and the control vector AAV/dSaCas9 with no repressor were administered by stereotaxic injection into the mouse dorsal hippocampus (DH) and validated using the mouse endogenous Apoe gene (a, b) AAV/gRNA(Apoe)p-dSaCas9-KRAB-MeCP2(TRD) vectors with gRNA1 or gRNA2 were injected into the right DH and the control AAV/dSaCas9 into the left DH. Both AAV/gRNA (Apoe)p-dSaCas9-KRAB-MeCP2(TRD) vectors reduced the mouse endogenous ApoE expression. a Representative images of brain coronal slices at 2× magnification 42 days post-injection, showing ApoE expression and DAPI staining in the DH; 20× magnification of DH region showing ApoE expression. b Signals were quantified using ImageJ. Box plot displays the ratios of the right DH relative to the left DH for mice injected with the repressor vector harboring gRNA1 (n = 8, p = 0.0002) and gRNA2 (n = 8, p = 0.0011). Each open circle represents the quantified signal right/left for a mouse. Values represent mean ± SEM. *p < 0.05 **p < 0.01 ***p < 0.001; Two -tailed Mann–Whitney U-Test. Source data are provided as a Source Data file.