Adeno-associated viral (AAV) vectors do not efficiently target muscle satellite cells

Abstract

Adeno-associated viral (AAV) vectors are becoming an important tool for gene therapy of numerous genetic and other disorders. Several recombinant AAV vectors (rAAV) have the ability to transduce striated muscles in a variety of animals following intramuscular and intravascular administration, and have attracted widespread interest for therapy of muscle disorders such as the muscular dystrophies. However, most studies have focused on the ability to transduce mature muscle cells, and have not examined the ability to target myogenic stem cells such as skeletal muscle satellite cells. Here we examined the relative ability of rAAV vectors derived from AAV6 to target myoblasts, myocytes and myotubes in culture and satellite cells and myofibers in vivo. AAV vectors are able to transduce proliferating myoblasts in culture, albeit with reduced efficiency relative to post-mitotic myocytes and myotubes. In contrast, quiescent satellite cells are refractory to transduction in adult mice. These results suggest that while muscle disorders characterized by myofiber regeneration can be slowed or halted by AAV transduction, little if any vector transduction can be obtained in myogenic stems cells that might other wise support ongoing muscle regeneration.

Figure 1

rAAV6 transduction of myotubes is more efficient that myoblast transduction in vitro. MM14 myoblasts or myotubes were transduced with rAAV6-CMV-AP, and both AP expression (a) and vector genome number (b) were quantified. Myoblasts were incubated with vector for either 1 hour (0d A) or 3 days (0d B) subsequent to withdrawal of FGF-2 and serum to halt proliferation. Shortened exposure to vector did not result in a statistically significant decrease in transduction. Myotubes were transduced at either 5 days (5d) or 10 days (10d) post-differentiation. Myotubes transduction at either timepoint was significantly increased compared to myoblasts. * indicates significantly different compared to 5d and 10d cohorts. ** indicates significantly different in comparison to 10 day cohort only. P < 0.05. FGF, fibroblast growth factor.

Figure 2

Vector-mediated reporter expression is diminished under conditions of high myoblast proliferation and turnover. C2C12 myoblasts were transduced with rAAV6-CMV-βGal at an MOI of 100. Cells were serially passaged every 2 days and diluted 1:10 with each passage. 90% of the population was discarded with each passage. At each passage, a cohort of wells (n = 2) was stained for βGal expression and the number of positive cells was quantified. Myoblasts gradually lost reporter expression over multiple cycles of dilution and proliferation. After four passages, the number of positive cells was reduced to ~2 per well. MOI, modality of infection.

Figure 3

In vitro rAAV6-mediated transduction of single myofiber cultures reveals a marked preference for multinucleated muscle cells. Single myofibers from both wt (top) or mdx (bottom) mice were isolated and cultured to establish a mixed population of mononuclear muscle progenitor cells and multinuclear myotubes. Cultures were transduced with rAAV6-mCherry on either day 7 or day 9. Cells were stained for desmin expression 2 days post-transduction, and the number of cells coexpressing both desmin and mCherry was quantified. Multinucleated muscle cells (dark bars) were preferentially targeted at both transduction timepoints relative to mononuclear cells (white bars). In cultures transduced on day 9, the total number of desmin-positive cells was reduced, but the proportion of myofiber transduction increased. *P < 0.05; ***P < 0.01; ****P < 0.001.

Figure 4

rAAV6 transduction efficiency is reduced during the early proliferative and inflammatory phase of muscle regeneration. Mice (n = 3) received an intramuscular (IM) injection of notexin in the extensor digitorus longus muscle, followed by rAAV6-CMV-AP injection at 2, 3, 4, or 5 days postinjury. Muscles were collected 2 weeks following rAAV6 injection and evaluated for AP expression (a) and vector genomes (b). We observed a trend towards increased transduction efficiency that correlated with an increase in the interval between injury and rAAV6 injection. Vector genome ratio is reported relative to copy number at the earliest injection timepoint (2 days).

Figure 5

Single fiber cultures from nestin-GFP mice injected with rAAV6. Mice received an intramuscular (IM) injection of rAAV6-CMV-mCherry in the extensor digitorus longus muscle. Fibers were isolated from injected muscles 4 weeks post-transduction and cultured for up to 6 days. At 24 hours, nestin/eGFP-positive satellite cells could be visualized on myofibers, but no cells stained for both eGFP and mCherry (a). eGFP-positive, mCherry-negative satellite cells were tracked as they migrated from the parent fibers and established myogenic colonies (b). Satellite cells remained negative for mCherry expression. By 6 days, numerous myotubes had formed in the cultures from the proliferating and terminally differentiating myoblasts (c), but the only eGFP-positive myotubes observed were those that had fused with the original myofiber (c, inset). eGFP, enhanced green fluorescent protein.

Figure 6

Analysis of satellite cell transduction in muscles injected with rAAV6. (a) Cross-sections of wild-type muscles injected with rAAV6-CMV-eGFP (a). Injected muscles were collected 4 weeks postinjection and stained for Pax7 and DAPI, then visualized under a fluorescence microscope. eGFP expression was not observed in any Pax7-positive cells. (b,c) Both wt and mdx mice were injected with rAAV6-CMV-mCherry with the dose indicated in (c). The high level of transduction using the 1 × 10¹⁰ vg dose was evident upon harvesting the muscles (b) (images taken at 7× magnification). No vector genomes were detected among satellite cells of either mouse strain 4 weeks postinjection (c). However, vector genomes were detected in DNA extracted from whole, injected extensor digitorus longus muscles (c). DAPI, 4′,6-diamidino-2-phenylindole; eGFP, enhanced green fluorescent protein.

Figure 7

Comparison of rAAV6, 8, and 9 for satellite cell transduction. Eight-week-old C57Bl/6 mice were injected intramuscularly into the TA muscle with rAAV-CMV-eGFP vectors pseudotyped with either the serotype 6 (left panels), 8 (middle panels), or 9 (right panels) capsids. After 2 weeks, muscles were harvested, and cryosections were imaged for eGFP (top panels). The inset shows background eGFP fluorescence in uninjected muscles. Cryosections were also stained with DAPI, and immunostained with antibodies against laminin (not shown) and Pax7. Vector doses were adjusted to generate similar levels of transduction (rAAV6: 5 × 10¹⁰ vg; rAAV8 and 9: 2 × 10¹² vg; N = 4 muscles each). As in the studies in Figure 6, no Pax7⁺/eGFP⁺ cells were observed following rAAV6 injection. However, a few rare instances were observed of cells between the basal lamina and the sarcolemma that were Pax7-negative/eGFP+ (bottom left panel, arrowheads). As with rAAV6, no Pax7⁺/eGFP⁺ cells were observed following rAAV9 injection (e.g., bottom right panel, arrowhead points to a Pax7⁺ satellite cell). In contrast, following injection of rAAV8 several Pax7⁺/eGFP⁺ cells were observed, but these represented less than 5% of the total Pax7⁺ cells scored (bottom of middle panel, arrowhead; the positive satellite cell adjacent to the myofiber with the asterisk is shown in the inset minus the red (Pax7) fluorescence). Scale bars: 150 µm top; 30 µm bottom. DAPI, 4′,6-diamidino-2-phenylindole; eGFP, enhanced green fluorescent protein.

Figure 8

Analysis of satellite cell transduction following systemic injection of rAAV6. Four-week-old C57Bl/6 mice were retro-orbitally injected with 4 × 10¹² vg of rAAV6-CMV-eGFP and analyzed 4 weeks later (N = 4 mice). (a) eGFP fluorescence in a cryosection of an isolated TA muscle showing transduced myofibers. (b–e) cryosections were stained with DAPI to visualize nuclei, and immunostainied with laminin and Pax7. An example of a Pax7⁺/eGFP-negative satellite cell is shown. Laminin + DAPI imaging (b); Pax7 + DAPI + eGFP imaging (c); GFP + DAPI imaging (d); merge (e). The arrowhead points to a Pax7⁺/eGFP-negative satellite cell. No Pax7⁺/eGFP⁺ cells were observed. Scale bars: 150 µm. DAPI, 4′,6-diamidino-2-phenylindole; eGFP, enhanced green fluorescent protein.