AAV-Mediated Targeting of the Activin A-ACVR1R206H Signaling in Fibrodysplasia Ossificans Progressiva

Abstract

Fibrodysplasia ossificans progressiva (FOP) is an ultra-rare genetic disorder characterized by progressive disabling heterotopic ossification (HO) at extra-skeletal sites. Here, we developed adeno-associated virus (AAV)-based gene therapy that suppresses trauma-induced HO in FOP mice harboring a heterozygous allele of human ACVR1^R206H (Acvr1^R206H/+) while limiting the expression in non-skeletal organs such as the brain, heart, lung, liver, and kidney. AAV gene therapy carrying the combination of codon-optimized human ACVR1 (ACVR1^opt) and artificial miRNAs targeting Activin A and its receptor ACVR1^R206H ablated the aberrant activation of BMP-Smad1/5 signaling and the osteogenic differentiation of Acvr1^R206H/+ skeletal progenitors. The local delivery of AAV gene therapy to HO-causing cells in the skeletal muscle resulted in a significant decrease in endochondral bone formation in Acvr1^R206H/+ mice. These mice showed little to no expression in a major AAV-targeted organ, the liver, due to liver-abundant miR-122-mediated repression. Thus, AAV gene therapy is a promising therapeutic strategy to explore in suppressing HO in FOP.

Keywords: AAV; ACVR1; Activin A; fibrodysplasia ossificans progressiva; gene therapy; heterotopic ossificans.

Figure 1

Elevated levels of Activin A in HO tissues of FOP mice. (A) Pinch injury was introduced into the right-side of the gastrocnemius muscle of the 8-week-old Acvr1^R206H/+ mice, and four weeks later, trauma-induced HO was assessed via radiography (left). Alternatively, the protein levels of Activin A in tissue lysates were measured using ELISA (right, n = 3). HB: heterotopic bone. (B) Protein levels of Activin A in the supernatant of mouse Acvr1^+/+;Prx1 (WT) and Acvr1^R206H;Prx1 (FOP) BMSCs (n = 4). (C,D) Mouse (C, n = 4) or human (D, n = 3) BMSCs were cultured under non-differentiation (ND) or osteoblast differentiation (OBD) conditions in the presence of PBS, LPS, TNF, or IL-1β, and Activin A expression in the supernatant was measured via ELISA. (E) Mouse Acvr1^R206H;Prx1 BMSCs were transduced using AAV9 carrying control (amiR-ctrl) or amiR-ACVR1^R206H.ACVR1^opt and then cultured under osteogenic conditions in the presence of PBS, LPS, TNF, or IL-1β. Twenty-four hours later, Activin A expression in the supernatant was measured by ELISA (n = 3). **, p < 0.01; ***, p < 0.001, and ****, p < 0.0001 by an unpaired, two-tailed Student’s t-test (A–C) and one-way ANOVA test (D,E).

Figure 2

Generation of an Activin A-silencing AAV vector. (A) HA-tagged mouse or human Inhba plasmid was transfected into HEK293 cells along with the control (amiR-ctrl) or amiR-Inhba #1–#6 plasmid and then total cell lysates were immunoblotted with the HA antibody. Gapdh was used as a loading control. amiR-Inhba #1–#6 indicate six amiRs targeting the shared coding sequences of mouse and human Inhba. Original images can be found in Figure S3. (B) Mouse Acvr1^R206H;Prx1 BMSCs or human BMSCs were transduced via AAV9 carrying amiR-ctrl or amiR-Inhba #4, and mRNA levels of Inhba were assessed using RT-PCR (n = 4). (C) AAV-transduced mouse Acvr1^R206H;Prx1 BMSCs or human BMSCs were cultured in the presence of PBS, LPS, TNF, or IL-1β, and 24 h later, Activin A expression in the supernatant was measured using ELISA (n = 3). *, p < 0.05; **, p < 0.01; and ***, p < 0.001 via an unpaired two-tailed Student’s t-test (B) and one-way ANOVA test (C).

Figure 3

Biodistribution of AAV9.gfp.MIR in mice. A total of 5 × 10¹² vg/kg of AAV9.gfp.MIR was transdermally (t.d.) injected into 8-week-old Acvr1^R206H/+ mice (n = 3), and 2 weeks later, individual tissue distribution of AAVs was assessed via EGFP expression using the IVIS optical imaging system (A) or histology on a frozen section of AAV-treated tissues (B). Scale bars: 100 μm.

Figure 4

AAV gene therapy suppresses Activin A signaling and osteogenic differentiation of FOP skeletal progenitors. (A) A schematic diagram showing AAV vector genome and capsid. AAV vector genome containing the CBA promoter, amiR-ACVR1^R026H, amiR-Inhba #4 (red), codon-optimized human ACVR1 (ACVR1^opt), and liver-abundant miR-122 target sequences (TS) was packaged into AAV9 capsid. CBA: chicken β actin; PA: poly A sequences. (B–D) Mouse Acvr1^R206H;Prx1 BMSCs were transduced by AAV9 carrying gfp.MIR (ctrl) or Acvr1/Inhba.MIR and cultured under osteogenic conditions. Six days later, gene expression was measured via RT-PCR and normalized to Gapdh (B,D, n = 4). Osteogenic differentiation was assessed by ALP activity (D, n = 5). (E–G) AAV-transduced BMSCs were cultured under osteogenic conditions in the presence of PBS or Activin A (50 ng/mL). ALP activity (3 day culture) or alizarin red staining (10 day culture) were performed for early or late osteogenic differentiation, respectively (D, n = 5). Activin A-induced expression of Id1 and Msx2 was assessed via RT-PCR 24 h post-stimulation (E, n = 4). Cells were stimulated with Activin A at different time points and immunoblotted for phospho-Smad1/5. Hsp90 was used as a loading control (F). Original images of (G) can be found in Figure S4. **, p < 0.01; ***, p < 0.001, and ****, p < 0.0001 with an unpaired two-tailed Student’s t-test (A–C) and one-way ANOVA test (D,E).

Figure 5

Local delivery of AAV gene therapy prevents trauma-induced HO in FOP mice. (A) Diagram of study and treatment methods. A total of 5 × 10¹² vg/kg of AAV9 carrying gfp.MIR (ctrl) or Acvr1/Inhba.MIR was injected t.d. into the gastrocnemius muscle of 8-week-old Acvr1^R206H/+;Pdgfrα-GFP mice 3 days prior to pinch injury. Four weeks later, ACVR1^R206H, ACVR1^OPT, and Inhba expression was assessed using RT-PCR (B, n = 4), and HO in the injured muscle was assessed via X-ray (C), microCT (D), and histology on a frozen section of AAV-treated hindlimbs (E). Three-dimensional reconstruction images (D, left) and quantification of HO volume (D, right, n = 7~10) are displayed. Red arrow indicates heterotopic bone (C). Proliferation of Pdgfrα-GFP-expressing FAPs within HO tissues was visualized using fluorescence microscopy. Scale bars: 100 μm. HB: heterotopic bone; M: muscle. The frequency of GFP⁺ScaI⁺CD31⁻CD45⁻ FAPs in the injured muscle was assessed using flow cytometry (F), and Pdgfrα-GFP FAPs were FACS-sorted from the injured muscle and subjected to RT-PCR analysis (G, n = 4). **, p < 0.01 and ****, p < 0.0001 using an unpaired, two-tailed Student’s t-test (B,D,F,G).

Figure 6

Local delivery of AAV gene therapy suppresses progression of traumatic HO in FOP mice. (A) Diagram of study and treatment methods. After pinch injury on the gastrocnemius muscle of 8-week-old Acvr1^R206H/+ mice at day 0, t.d. injections of AAV9.Acvr1/Inhba.MIR (5 × 10¹² vg/kg) into the injured muscle were performed on days 1, 3, or 6 post-injury. AAV9.gfp.MIR (ctrl) was t.d. injected on the same day of pinch injury. ACVR1^R206H, ACVR1^OPT, and Inhba expression were assessed via RT-PCR at 4 weeks post-injury (B, n = 4). HO in the injured muscle was assessed using X-ray (C, top), microCT (C, bottom), and histology on a paraffin section of AAV-treated hindlimbs (D). Three-dimensional reconstruction images (C) and quantification of HO volume (D, n = 9~10) are displayed. Red arrows indicate heterotopic bones (C). Longitudinal sections of the injured muscle was stained with Alcian blue/hematoxylin/orange G (E). HB: heterotopic bone; F: fibrotic tissue; C: chondrogenic analgen; M: skeletal muscle. Scale bars: 100 μm. ***, p < 0.001 ****, p < 0.0001 using a one-way ANOVA test (A,B,D).