Systemic delivery of genes to striated muscles using adeno-associated viral vectors

Abstract

A major obstacle limiting gene therapy for diseases of the heart and skeletal muscles is an inability to deliver genes systemically to muscles of an adult organism. Systemic gene transfer to striated muscles is hampered by the vascular endothelium, which represents a barrier to distribution of vectors via the circulation. Here we show the first evidence of widespread transduction of both cardiac and skeletal muscles in an adult mammal, after a single intravenous administration of recombinant adeno-associated virus pseudotype 6 vectors. The inclusion of vascular endothelium growth factor/vascular permeability factor, to achieve acute permeabilization of the peripheral microvasculature, enhanced tissue transduction at lower vector doses. This technique enabled widespread muscle-specific expression of a functional micro-dystrophin in the skeletal muscles of dystrophin-deficient mdx mice, which model Duchenne muscular dystrophy. We propose that these methods may be applicable for systemic delivery of a wide variety of genes to the striated muscles of adult mammals.

Figure 1

Muscle transduction (blue) following systemic rAAV6 administration. (a–e) Mice were examined at 11 (a–c) or 14 (d,e) days after administration of 2 × 10¹¹ (a,b), 1 × 10¹² (c, d) or 1 × 10¹³ (e) vector genomes of rAAV6–CMV–lacZ (a–c) or rAAV6–CK6–lacZ (d,e) alone (a) or with 10 μg VEGF (b–e).

Figure 2

β-Gal activity in muscles after systemic rAAV6 administration. Bars represent activity (RLU: relative light units) from mice administered: (a) 2 × 10¹¹ vector genomes rAAV6–CMV–lacZ (gray, only soleus above background P ≤0.05 with Student’s t-test) or vector with 10 μg of VEGF (red); (b) 1 × 10¹² vector genomes of rAAV6–CMV–lacZ (black) or vector and 10 μg VEGF (dark red); (c) 1 × 10¹² vector genomes (light blue) or 1 × 10¹³ vector genomes (dark blue) of rAAV6–CK6–lacZ, both with 10 μg VEGF. Data from uninjected animals (white bars) are shown in a. Data represent means ± s.e. Ma, masseter; FL, flexor digitorum profundus/carpi radialis; Bi, biceps; Tr, triceps; Ht, heart; Di, diaphragm; Qd, quadriceps; TA, tibialis anterior; Ga, gastrocnemius; So, soleus. (d), Anti-β-gal immunofluorescence microscopy of the quadriceps muscles: a mouse administered 1 × 10¹² vector genomes of rAAV6–CMV–lacZ by intravenous injection (d and inset right), compared with an uninjected mouse (d, inset left). Animals examined 11 (a,b,d) or 14 (c) days post-treatment. Bars, 1 mm (d) and 100 μm.

Figure 3

β-Gal activity and vector genome copy numbers in muscles and organs after systemic administration of rAAV6 vectors with VEGF. (a,b) β-gal activity (a) and rAAV genome copy number (b) in striated muscles after intravenous injection of 2 × 10¹¹ vector genomes of rAAV6–CMV–lacZ coadministered with 1 (yellow; genomes not determined for TA), 5 (orange), or 10 (red) μg of VEGF, compared with no VEGF administration (gray); 1 arbitrary unit = 8 × 10⁴ vector genomes μg/DNA. (c) β-gal activity in a variety of organs after intravenous injection of 1 × 10¹² vector genomes of rAAV6–CMV–lacZ alone (black) or with 10 μg of VEGF (dark red) (d) rAAV6 genomes are detectable in organs from c; 1 arbitrary unit = 9 × 10⁶ vector genomes/μg DNA. Data from uninjected animals (white bars) are shown in a and c. Data are mean values ± s.e. TA, tibialis anterior; Br, brain; Lu, lung; Li, liver; Sp, spleen; In, intestine; Ki, kidney; Te, testes. Animals were examined 11 days post-treatment.

Figure 4

Myocardial transduction (blue) in adult mice 11 days after systemic administration of 1 × 10¹² vector genomes of rAAV6–CMV–lacZ and 10 μg VEGF. (a) The heart of a treated wild-type mouse examined after reaction with X-gal and counterstaining with hematoxylin and eosin. (b–i) Sections examined following reaction with X-gal and counterstaining with hematoxylin and eosin (b,d,f,h) and serial sections immunofluorescence-labeled for CD4-positive cells (red; c,e,g,i). The hearts of an untreated mouse (b,c) and a mouse administered empty vector capsids with VEGF (d,e) display regular morphology and few CD4-postive cells. (f,g) Regions of myocardium in the heart of a wild-type mouse administered 1 × 10¹² vector genomes of rAAV6–CMV–lacZ and VEGF show dramatic mononuclear cell and CD4-postive cell infiltration. (h,i) Myocardium of a treated, β-gal-tolerant transgenic mouse (villin mouse19), demonstrating a lack of cellular infiltrate. Scale bars, 1 mm (a) and 100 μm (b–i).

Figure 5

Systemic delivery of microdystrophin to dystrophic mice. (a) Anti-dystrophin immunofluorescence microscopy of tibialis anterior muscles from treated mdx mice (Tmdx) administered 1 × 10¹² vector genomes of rAAV6–CK6–microdystrophin and 10 μg VEGF, compared with wild-type and untreated controls. Dystrophin expression is increased in the muscles of treated compared with untreated mdx mice, but remains mosaic compared with wild-type mice. (b) Force producing capacity (P_o, top) and deficit (bottom), of tibialis anterior muscles from Tmdx mice (red) compared with wild-type (black) and untreated mdx (gray) mice after consecutive eccentric contractions (LC1, LC2, respectively). P values with Student’s t-test between Tmdx and mdx: i = 0.078, ii ≤0.05, iii = 0.074 and iv ≤0.05. (c) Anti-dystrophin-labeled muscles from mice administered 1 × 10¹³ vector genomes of rAAV6–CK6–microdystrophin and VEGF. The dystrophin expression pattern observed at this vector dose is no longer mosaic, in contrast to a. Mice were examined at 8 weeks (a,b) and 6 weeks (c) after treatment. Scale bars, 100 μm.

Figure 6

Sustained expression of a therapeutic human microdystrophin protein is achieved after intravenous administration of rAAV6 vectors, without significant cellular infiltration. (a) Anti-dystrophin-labeled tibialis anterior from a treated mdx mouse administered 1 × 10¹³ vector genomes rAAV6–CK6–microdystrophin and 10 μg VEGF. This treatment significantly (Tmdx compared with mdx, P ≤0.05) reduced serum creatine kinase (CK) levels. Individual data are depicted as dots with group means in red. (b) Anti-dystrophin immunofluorescence microscopy (left) and hematoxylin-stained and eosin-stained sections (right) of tibialis anterior muscles from wild-type, untreated mdx (mdx) and mdx mice treated as in a (Tmdx), respectively. Yellow points function are markers identifying the same muscle fiber in left and right panels. Mice were examined 6 weeks after treatment. Scale bars, 1 mm (a) and 200 μm (b).