Introduction

Adeno-associated viruses (AAV) belong to the viral family Parvoviridae (parvoviruses). The Parvoviridae infect a diverse number of vertebrates, including humans¹. AAV has a nonenveloped proteinaceous capsid with a diameter of approximately 20 to 25 nm². The capsid is composed of 60 protein subunits of three viral proteins (VP1, VP2, and VP3 from the Cap ORF) and encapsidates a linear, single-stranded DNA genome of approximately 4.7 kb^3,4,5. The genome is capped by inverted terminal repeats (ITRs) of roughly 150 bp^6,7,8, which form secondary structure necessary for stability and replication and are a necessary and sufficient packaging signal⁹. In addition to capsid proteins, the AAV genome codes for various Rep isoforms, proteins involved in replication and latent infection¹⁰. AAV is a dependovirus, so named because members of this group cannot replicate autonomously. AAV requires co-infection with a helper virus, such as herpesvirus or adenovirus, providing necessary helper functions in trans to allow for AAV replication^11,12,13. In the absence of a helper virus, AAV can undergo latent infection as an episome or it may integrate into the host genome^14,15. Numerous serotypes of AAV have been identified^{16,17,18,19,20,21}. Some serotypes are highly similar (AAV1 and AAV6)^16,17, while others show considerable divergence (AAV2 and AAV5)²². The various serotypes have a variety of tissue tropisms and binding characteristics and have been shown to infect many tissues efficiently, including the liver²³, lung²⁴, and striated musculature^25,26. Importantly, no AAV serotype has been identified as the causative agent of any human pathology²⁷.

The broad tissue tropism, efficient infection, and lack of human pathology displayed by AAV have led to its exploitation as a gene transfer vector, and recombinant AAVs (rAAVs) have been developed as potential gene therapy vectors²⁸. These rAAV vectors lack all viral genome sequence except the ITRs, between which expression cassettes can be inserted. These recombinant genomes can be efficiently packaged provided their size is not significantly larger than that of the wild-type genome, though they can be considerably smaller²⁹. To be packaged into functional vectors, these recombinant genomes must be provided with all Rep/Cap and helper functions in trans. This technique has been refined into a two- or three-plasmid cotransfection protocol in which all helper functions are carried on plasmids, removing the possibility of helper virus contamination of the purified rAAV stock^28,30. Depending on the serotype, rAAV vectors can be purified through various methods such as density centrifugation or affinity chromatography^16,24,28. Conveniently, the capsid proteins from most serotypes can cross-package with the ITR sequences of alternate serotypes, enabling easy and rapid generation of pseudotyped vectors carrying ITRs from one serotype packaged by capsids from an alternate serotype^23,31.

Most gene therapy applications have utilized rAAV2-based vectors for gene transfer. AAV2 has been shown to transduce numerous tissues without elicitation of a cellular immune response and can lead to stable gene expression up to several years^32,33,34. However, transduction of skeletal muscles by rAAV2 vectors has been reported to be influenced by fiber type, such that slow-twitch fibers are more readily transduced than are fast-twitch fibers³⁵. The skeletal musculature is severely affected in many disorders and develops increasing weakness with advancing age^36,37. Gene therapy for many muscle disorders, such as Duchenne muscular dystrophy, could benefit by the availability of an AAV vector able to transduce all muscle types more efficiently. Here we have characterized skeletal muscle transduction with vectors pseudotyped with the serotype 6 capsid¹⁷. rAAV6-pseudotyped vectors have previously shown greatly increased transduction of both lung and liver compared with rAAV2 vectors^23,24. We demonstrate that rAAV6-pseudotyped vectors can transduce the skeletal musculature at levels >500-fold higher than rAAV2 vectors. These vectors also reach peak expression levels earlier than rAAV2 vectors. These findings demonstrate that rAAV6-pseudotyped vectors are an extremely efficient tool for genetic transfer to the skeletal musculature and may prove useful in therapeutic gene delivery protocols.

Results and discussion

rAAV6 production and purification

Our rAAV production and purification method is based on a previously reported technique using calcium phosphate-mediated transfection and heparin-affinity chromatography²⁴. A representative purification tracing of OD₂₈₀ and conductivity is shown (Fig. 1A). This approach resulted in purified virus, free of protein impurities as detected by Coomassie blue staining of an SDS–PAGE gel. The Coomassie-stained gel demonstrates detectable presence of only the three viral capsids, VP1, VP2, and VP3, when 5 10⁹ or 4 10¹⁰ vector genomes (vg) were analyzed (Fig. 1B). This vector production was efficient, requiring 3 days from transfection, and did not require compounds to boost transfection efficiency nor medium adjustment dependent on metabolite levels³⁸. The production protocol used here generated typical yields of 2 10¹³ vg of rAAV6 at a concentration of 5 10¹² vg/ml from 2 10⁸ HEK293 cells. This corresponds to a yield approaching 10⁵ vg per cell. We obtained similar yields with rAAV2 vectors using a slightly modified technique as outlined under Methods. Our studies were normalized to vector genomes, as we observed a wide range of genome-to-infectivity ratios in vitro that did not correspond to in vivo transduction levels. In HT1080 cells, we observed a genome-to-infectivity ratio of 10² to 10³:1 for rAAV2 and 10⁵ to 10⁶:1 for rAAV6 when we used a CMV-lacZ expression cassette¹⁷.

Figure 1.

Purification and concentration of rAAV6 vectors. (A) A representative tracing from an rAAV heparin-affinity purification with OD₂₈₀ and conductivity in microsiemens (s) shown. The loading, wash, and elution stages are denoted L, W, and E, respectively. The detergent eluted peak is noted with an asterisk. (B) An SDS–PAGE gel of samples of rAAV6 vectors collected after heparin column purification stained with Coomassie blue for viral capsid protein detection (VP1, VP2, and VP3). The gel was loaded with either 5 10⁹ or 4 10¹⁰ vector genomes (vg) of material from rAAV2 and rAAV6 vectors expressing either microdystrophin or galactosidase.

Full figure and legend (164K)

Transduction of skeletal muscle using rAAV6

While rAAV2-based vectors have been reported to transduce many tissues, including the skeletal musculature²⁶, vectors based on some alternate serotypes have proved more efficient^23,39. Though not in common use, vectors based on rAAV6 have been shown to yield drastically greater levels of transduction in tissues such as lung compared with rAAV2 vectors²⁴. To compare the efficiency and extent of muscle transduction achievable via rAAV6 and rAAV2 vectors, we injected rAAV6 and rAAV2 vectors encoding -gal (rAAV2:CMV-lacZ and rAAV6:CMV-lacZ) directly into hindlimb muscles of mice. As rAAV2 vectors have been reported to transduce preferentially slow-twitch muscle fibers, we examined the soleus and extensor digitorum longus (EDL) muscles to include a significant population of slow- and fast-twitch fibers, respectively³⁵.

When we injected rAAV6:CMV-lacZ at a dose of 2 10⁹ vg, the EDL and soleus muscles expressed high and uniform levels of -gal from 2 weeks out to at least 12 weeks (Fig. 2). An identical dose of rAAV2 vector resulted in markedly lower levels of transduction (especially in the EDL). Reporter gene expression from rAAV2 was more mosaic, and delayed in expression by 2 to 4 weeks, in both the soleus and the EDL muscles compared with that from rAAV6 (Figs. 2 and 3). At this dose, rAAV6 vectors displayed levels of transduction at least 50-fold (P 0.05) greater than were obtained with rAAV2 vectors in the soleus and more than 500-fold (P 0.05) greater than in the EDL (Fig. 3). At lower vector doses, transduction levels were considerably reduced and more variable (Fig. 3). At these doses (2 10⁵ and 2 10⁷), we have observed a threshold effect by which the same dose of virus can yield variable transduction results from injection to injection, possibly due to low vector doses exacerbating inherent variability, e.g., possible nonspecific binding to the syringe or connective tissue or minor unobserved leakage of vector injectate. At 2 10⁷ vg per muscle, rAAV6 vectors yielded statistically significant transduction in all but the 4-week soleus samples, while the rAAV2 vectors demonstrated significant (P 0.05) transduction in only the 2-week EDL samples (Fig. 3). Neither vector yielded significant (P 0.05) transduction in either muscle at any time point studied at the 2 10⁵ vg per muscle dose (Fig. 3). These results demonstrated that rAAV6 vectors lead to greater transduction of skeletal muscles than rAAV2 vectors at all effective doses tested and that both vectors have a minimum effective dose. These results are derived from four to six muscles per variable.

Figure 2.

Persistent transduction of murine skeletal muscles by rAAV6-pseudotyped vectors compared with rAAV2 vectors. Histochemical reaction for -galactosidase activity (blue) of transverse cryosections obtained from the extensor digitorum longus (EDL) and soleus muscles of mice injected with 2 10⁹ vg of rAAV2:CMV-lacZ or rAAV6:CMV-lacZ vectors 2, 4, 8, or 12 weeks before examination. Sections are counterstained with eosin. Representative examples from n = 4 to 6 individual muscles. Scale bar, 500 m.

Full figure and legend (200K)

Figure 3.

rAAV6-pseudotyped vectors demonstrate increased expression in mouse limb muscles compared with rAAV2 vectors. Mean -galactosidase activity of (a) extensor digitorum longus (EDL, a) and soleus (b) hindlimb muscles of untreated mice (white bars) and mice administered intramuscular injections of 2 10⁵ (light grey), 2 10⁷ (grey), or 2 10⁹ (black) vg of rAAV2:CMV-lacZ (columns labled 2) or rAAV6:CMV-lacZ (column labled 6) vectors examined 2 or 4 weeks after injection. Mean -galactosidase activity of EDL (c) and soleus (d) muscles of mice examined 2, 4, or 8 weeks after intramuscular injections of 2 10⁹ vg of rAAV2:CMV-lacZ (grey) or rAAV6:CMV-lacZ (black) vectors. Mean activity at time 0 represents samples of respective uninjected muscles.

Full figure and legend (203K)

These results highlight a significant advantage of rAAV6 over rAAV2 vectors for skeletal muscle transduction: one can use either the same dose of rAAV6 vector to achieve greater total transduction levels than would be obtained with rAAV2 or a lower dose of rAAV6 to achieve transduction levels comparable to a higher dose of rAAV2 vector. This advantage may allow a reduced vg/kg dose to be used to achieve a given transduction level, potentially increasing safety. Practically speaking, this efficiency also reduces the amount of vector required, enabling easier scale-up for either experimental or clinical use. The relatively rapid onset of expression observed with rAAV6 vectors could also translate into an important advantage when transgene expression is needed quickly, such as for manipulation of cell populations with short half-lives. Another example might be therapeutic delivery to the musculature in various dystrophies. Here, the target cell population undergoes continuous cycles of degeneration and regeneration, such that a transduced myofiber would need to express the exogenous protein before the next degeneration cycle or the vector would be cleared without therapeutic benefit. The onset kinetics and efficiency of transduction observed, coupled with the relative ease of production and purification, suggest that rAAV6 vectors may be extremely useful tools for skeletal muscle transduction.

We next sought to determine whether rAAV6 vectors displayed any fiber-type-specific transduction patterns. When transduction was measured on a whole-muscle activity level following injection of 1 10⁹ genomes, rAAV6 vectors transduced the EDL slightly more efficiently than the soleus, though not significantly so (P = 0.12, Fig. 3), demonstrating little fiber-type preference for this pseudotype. In contrast, rAAV2 vectors showed the opposite trend, with significantly greater transduction of soleus than EDL (P 0.05), in agreement with previous literature (Fig. 3)³⁵. As rAAV2 has been reported to transduce preferentially type I (slow-twitch) muscle fibers³⁵, we analyzed the soleus and EDL for transduction mosaicism based on fiber type using both rAAV6 and rAAV2 vectors. For this purpose, we stained serial muscle sections using histochemistry (data not shown) and immunofluorescence with antibodies against –galactosidase and types I, IIa, and IIb myosin heavy chain isoforms (Fig. 4 and see Methods). At the level of individual fibers, we saw no preference, suggesting that local factors, such as local viral concentration, may override any preferential transduction based on fiber type. We saw no clear correlation between -gal staining and fiber type either in areas of mosaic -gal staining or in areas containing multiple fiber types (Fig. 4). While our results support preferential transduction of slow-twitch fibers by whole-muscle analysis (Fig. 3), our analysis by immunofluorescence and histochemistry was unable to detect a similar phenomenon (Fig. 4). This result may reflect limitations in attempting quantitative measurements with these techniques. Thus, while the rAAV2 vector was clearly able to transduce all fiber types, type I fibers might take up more copies of the vector and therefore express greater levels of the transgene than do type II fibers. Alternatively, the transduction preference may be related to a characteristic of oxidative muscle that does not necessarily correlate with myosin isoforms. In either case, fiber-type preferences of these two serotypes are not absolute barriers to fiber transduction. Regardless, these results suggest another potential benefit of rAAV6 vectors, which is their ability to generate higher and more uniform levels of transduction in both fast- and slow-twitch muscles compared with rAAV2 vectors (Figs. 2,3–4).

Figure 4.

Mosaic reporter gene expression in murine muscle fibers administered rAAV vectors is not associated with phenotypic aspects related to distinct myosin heavy chain isoforms. Transverse cryosections of extensor digitorum longus (EDL; predominantly fast fiber type) and soleus (mixed fiber type) muscles of mice examined 8 weeks after intramuscular injection of 2 10⁹ vg of rAAV2:CMV-lacZ or rAAV6:CMV-lacZ vectors. Serial sections of each muscle were processed with immunofluorescence detection techniques specific to -galactosidase (-Gal) or myosin heavy chain isoform I, IIa, or IIb. Symbols identify a cell on serial sections within each muscle for orientation and comparison of serial sections. Scale bar, 200 m.

Full figure and legend (201K)

rAAV6 vectors are highly efficient at transducing respiratory muscles

Encouraged by the high and widespread transduction levels seen following rAAV6 vector delivery via direct muscular injection, we assessed the ability to transduce the diaphragm muscle of mice. The diaphragm muscle is of great interest as a target for gene therapy due to its central role in respiration, yet it has proven difficult to transduce due to its constant movement, relative thinness, and anatomical inaccessibility. Previous attempts to transduce the diaphragm with a variety of vectors have involved blind or visualized injection, circulatory isolation and vector administration, or surgery to expose the abdominal surface for "painting" on of a vector-containing gel and have met with varying degrees of success^{40,41,42,43,44}. We attempted to transduce the diaphragm via a simple and minimally invasive technique based on an injection of vector directly into the thoracic cavity. Injection of 5 10¹¹ vg in a total volume of 100 l resulted in widespread transduction of both the diaphragm and the intercostal muscles of mice when we utilized rAAV6 vectors (Fig. 5). Transduction was observed throughout the full thickness of the vast majority of the diaphragm musculature, though transduction was not observed in muscles outside of the thoracic cavity (Fig. 5 and data not shown). In contrast, rAAV2 vectors achieved drastically reduced transduction of the respiratory muscles compared with rAAV6 vectors. rAAV2-mediated transduction was absent in the intercostal muscles and transduction of the diaphragm was detectable only by histochemistry or immunofluorescence and was limited to superficially located myofibers (Fig. 5). The lower transduction seen with rAAV2 may be reflective of a reduced ability to penetrate the connective and musculature tissue due to specific or nonspecific binding or a relatively higher particle-to-infectivity ratio in vivo compared with rAAV6. We found this protocol to be highly reproducible (n = 10) and readily tolerated by all treated animals, with no adverse effects observed and a 100% survival rate (data not shown). These results represent the first demonstration of a technique capable of transducing the majority of the diaphragm and intercostal muscles of an adult animal without surgery or manipulation of the circulatory system.

Figure 5.

Intrathoracic administration of rAAV6 vectors to adult mice yields dramatic transduction of skeletal muscles compared with rAAV2 vectors. Gross anatomy of the forelimb and ribcage of an untreated adult mouse and mice examined 4 weeks after intrathoracic injection of 5 10¹¹ vg of rAAV2:CMV-lacZ or rAAV6:CMV-lacZ vector that have been histochemically reacted for -galactosidase activity (blue). Transverse cryosections of the diaphragm muscle were also examined for -galactosidase activity using histochemical methods (blue, counterstained with eosin) and immunofluorescent detection (green). Examples shown are representative of n = 5 for each group. Scale bar, 500 m.

Full figure and legend (311K)

rAAV6 vectors achieve body-wide transduction of the skeletal musculature following intraperitoneal injection into newborn mouse pups

Gene therapies for diseases of the skeletal musculature will require efficient gene delivery to muscles located throughout the body. Previous attempts to overcome this difficulty have resulted in impressive body-wide transduction of the skeletal musculature when vector is administered to mouse pups either in utero or shortly after birth^44,45,46. Presumably, this is due to the animal's small size and the incomplete formation of intracellular connections and connective tissue deposition in fetal and newborn mouse pups, allowing for relatively efficient vector dissemination. Since rAAV6 demonstrated efficient transduction of adult skeletal muscle, we examined the ability of rAAV6 vectors to achieve widespread transduction following injection into neonatal mice. We injected approximately 5 10¹¹ vg of rAAV6:CMV-lacZ in a total volume of 100 l into the intraperitoneal cavity of 1- to 3-day-old mouse pups. Four weeks postinjection we observed widespread transduction throughout the skeletal musculature. -Gal activity was greatest in the diaphragm and hindlimb musculature, with widespread but lower expression seen in the head, trunk, and forelimb musculature (Fig. 6 and data not shown, n 12 animals). This reduced expression may be attributable to greater physical barriers associated with dissemination to the more distal muscles, e.g., for dissemination to the forelimbs, the vector must pass the trunk, while the heart is separated from the intraperitoneal space by the ribcage, diaphragm, and lungs. This method was observed to be most efficient when applied to mice that were fewer than 5 days of age, and mice more than 10 days of age displayed greatly reduced transduction of skeletal muscles (data not shown).

Figure 6.

rAAV6-pseudotyped vectors achieve body-wide transgene expression after intraperitoneal injection in newborn (3 days of age) mice. Medial aspects of the fore- and hindlimbs and inferior surface of the diaphragm muscles of an uninjected mouse and a mouse examined 2 weeks after intraperitoneal injection of 1 10¹¹ vg rAAV6:CMV-lacZ at 3 days of age. Images in the right column show immunofluorescent signal detection of -galactosidase activity (green) in transverse cryosections of the diaphragm muscles of additional, similarly treated mice. Examples are representative of levels of transduction from 12 animals per individual serotype. Scale bar, 500 m.

Full figure and legend (118K)

Though this technique allows for "proof of principle" experiments for gene delivery to the skeletal musculature, it will likely be of limited therapeutic use, as this permissiveness is lost shortly after birth. Therefore, this approach may be useful only for interventions early in development. However, this technique is a potentially valuable research tool, allowing the creation of "pseudotransgenic" mice with respect to at least the musculature via vector administration early in life. This scenario would enable screening of expression cassettes more rapidly than with classic transgenic mouse construction. The generation of transgenic mice represents a substantial investment in time and resources, thus the ability to screen potential expression cassettes before committing to the construction of a transgenic mouse is extremely desirable.

rAAV is an efficient gene delivery vector that has a positive safety record in human clinical trials following delivery to muscle and lung^47,48,49. To date, only rAAV2 vectors have been used in clinical trials; however, alternative serotypes are garnering greater interest due to their increased tropism for various tissues^23,24,39. Here we demonstrated that vectors based on the rAAV6 pseudotype are extremely efficient at gene transfer to the skeletal musculature. rAAV6 vectors are readily produced at high titer and purity and transduce skeletal muscle much more efficiently than rAAV2 vectors. This highly efficient transduction allows for the near saturation of murine muscles following direct intramuscular injection. Further, rAAV6, but not rAAV2, enables rapid, reproducible, and efficient transduction of the diaphragm and intercostal muscles of mice via a simple and noninvasive intrathoracic injection. The ease and extent of transduction should greatly facilitate studies of these critical respiratory muscles. rAAV6 vectors also result in widespread transduction of the musculature following intraperitoneal injection into newborn mice. This characteristic is of high value, as it allows for the genetic manipulation of body-wide musculature and the screening of expression cassettes for therapeutic potential before transgenic mouse construction. Finally, we note that the increased tropism for skeletal muscle displayed by rAAV6 vectors may be exploited to achieve whole-body, systemic gene delivery to adult mice⁵⁰. Based on these observations, we conclude that rAAV6 vectors may contribute greatly to studies dependent on genetic transfer to the skeletal musculature and potentially to clinical gene therapy trials for diseases of the skeletal musculature.

Methods

Plasmids, viral production, and titering

Vectors were produced by CaPO₄ cotransfection of HEK293 cells with the plasmid pAAV-CMV-lacZ (Stratagene, La Jolla, CA, USA) with either pDG²⁸ or pA6DG⁵⁰ for production of rAAV2:CMV-lacZ and rAAV6:CMV-lacZ, respectively. The packaging plasmids pDG and pDGM6 are identical except that the Cap ORF for AAV6 was inserted as a SwaI/ClaI fragment into pDG in place of the AAV2 Cap ORF⁵⁰.

The purification method described here is a variation of previously reported techniques²⁴. Our modifications include homogenization using a 110S microfluidizer (Microfluidics, Newton, MA, USA) followed by clarification of the homogenate by filtration through a 0.22-m filter. Additionally, an Amersham AKTA10 HPLC machine (Piscataway, NJ, USA) was used for affinity purification on a HiTrap heparin column (Amersham). After the vector was loaded in clarified DMEM, the column was washed, in order, with Ringer's solution, Ringer's solution with 0.5% w/v N-lauroyl sarcosine, and Ringer's solution with 200 mM (300 mM for rAAV2) NaCl. The vector was then eluted in Ringer's solution with 400 mM (500 mM for rAAV2) NaCl and dialyzed against physiological Ringer's solution and frozen until use.

Vector concentration was titered by quantitative dot-blot analysis using the CDP-Star chemiluminescence kit (Amersham) and visualized with a GeneGnome Imager (Syngene, Frederick, MD, USA). Standards were the lacZ-containing EagI fragment from plasmid pAAV-CMV-lacZ. The probe used corresponded to internal lacZ fragments from a BssHII digest of pAAV-CMV-lacZ.

Experimental animal manipulation

All experimental manipulation of animals was conducted according to protocols approved by the Institutional Animal Care and Use Committee of the University of Washington. For intramuscular injections, animals were anesthetized with either 2,2,2-tribromoethanol (Sigma, St. Louis, MO, USA) or isoflurane (Abbott Laboratories, North Chicago, IL, USA). The skin of the lower leg was opened to allow access to the EDL and soleus muscles. Using a 32-gauge Hamilton syringe (Reno, NV, USA), muscles were injected with a total volume of 20 l of physiological Ringer's solution containing the appropriate amount of vector. The skin was then closed with Nexaband surgical grade adhesive (World Precision Instruments, Inc., Sarasota, FL, USA) and animals were monitored for recovery and wound healing until time of sacrifice. For intrathoracic injections, animals were anesthetized with isoflurane, and a 29-gauge insulin syringe (Becton–Dickinson, Franklin Lakes, NJ, USA) was used to inject 100 l of vector in physiological Ringer's solution between the ribs into the intrathoracic space, guided by visualization and palpation. For intraperitoneal injections, animals were anesthetized via cooling, and a 29-gauge insulin syringe was used to inject 100 l of vector in physiological Ringer's solution into the intraperitoneal space. Animals reported in this study were young adult (5–6 week) C57Bl/6J females obtained from The Jackson Laboratory (Bar Harbor, ME, USA).

Luminometer assays

After sacrifice, the relevant muscles were rapidly excised and flash frozen in liquid nitrogen. The frozen muscles were then powdered using a mortar and pestle and the protein was extracted with a protease-inhibiting buffer containing 137 mM NaCl, 20 mM Tris–HCl, pH 7.6, 2 mM MgCl₂, 1 mM 2-mercaptoethanol, 0.2% Tween 20 (Amersham), and 1 Complete protease inhibitor (Roche, Indianapolis, IN, USA). The extract was then measured for protein concentration via spectrophotometric absorption with the Bradford reagent (Pierce, Rockford, IL, USA). The extract was analyzed for -gal activity with a commercial luminometry kit from BD Biosciences (San Diego, CA, USA) according to the product manual; activity was read on a Sirius luminometer (Berthold, Oak Ridge, TN, USA). Endogenous -gal activity was inactivated by incubation at 50°C for 1 h. All protein concentration and -gal activity assays were averaged over a triplicate reading and represent n = 4–6 individual muscles.

Immunofluorescence and histological analysis

For muscle sections, the relevant muscles were rapidly excised after sacrifice, coated in Optimal Cutting Temperature compound (Sakura Finetek, Torrance, CA, USA), and frozen in liquid nitrogen-cooled isopentane. For histochemistry, 5-m sections were fixed for 10 min in pH 7.4 Sorenson's buffer supplemented with 0.5% glutaraldehyde for 10 min at room temperature. Sections were then rinsed with PBS and reacted at 37°C for 4 h in pH 7.4 Sorenson's buffer supplemented with 1 mg/ml X-gal. Sections were counterstained with eosin/phloxine. For immunofluorescence, 5-m sections were incubated with rabbit anti--galactosidase antibody from Molecular Probes (Eugene, OR, USA) or the myosin heavy chain isoform-specific monoclonal antibodies BA-D5 (type I), BF-F3 (type IIb), and SC-71 (type IIa) from ATCC (Manassas, VA, USA)⁵¹. The secondary goat anti-rabbit antibody, Alexa 488 (Molecular Probes), was used as the fluorescent label for the anti--gal antibody, while the M.O.M. kit from Vector Laboratories (Burlingame, CA, USA) was used to detect the myosin heavy chain antibodies. Bright-field and fluorescent images were taken on a Nikon E1000 microscope (Melville, NY, USA). For whole-mount tissues, animals were rapidly dissected after sacrifice, rinsed in PBS, and then fixed in pH 7.4 Sorenson's buffer supplemented with 4% paraformaldehyde at room temperature for 15 min under -760 Torr. The tissues were then rinsed in PBS and reacted at 37°C for 8 h under -760 Torr in pH 7.4 Sorenson's buffer supplemented with 1 mg/ml X-gal. After staining, tissues were rinsed in PBS and fixed for 48 h at 4°C in Karnovsky's fixative before transfer to 70% ethanol.

Statistical treatments

ANOVA was used to determine differences between groups for all comparisons. This includes differences between all groups and background signal, between groups injected with rAAV2 vs rAAV6, and within data sets injected with the same pseudotype, analyzed at different time points. ANOVA was performed, followed by post hoc analysis via the Scheffe and Bonferroni/Dunn methods. The maximum probability of a type I error was set at 0.05.

References

1. Siegl, G., et al. (1985). Characteristics and taxonomy of Parvoviridae. Intervirology. 23: 61–73. | PubMed | ChemPort |
2. Xie, Q., et al. (2002). The atomic structure of adeno-associated virus (AAV-2), a vector for human gene therapy. Proc. Natl. Acad. Sci. USA. 99: 10405–10410. | Article | PubMed | ChemPort |
3. Green, M. R. and Roeder, R. G. (1980). Definition of a novel promoter for the major adenovirus-associated virus mRNA. Cell. 22: 231–242. | Article | PubMed | ISI | ChemPort |
4. Berns, K. I. and Adler, S. (1972). Separation of two types of adeno-associated virus particles containing complementary polynucleotide chains. J. Virol. 9: 394–396. | PubMed | ChemPort |
5. Srivastava, A., Lusby, E. W. and Berns, K. I. (1983). Nucleotide sequence and organization of the adeno-associated virus 2 genome. J. Virol. 45: 555–564. | PubMed | ISI | ChemPort |
6. Straus, S. E., Sebring, E. D. and Rose, J. A. (1976). Concatemers of alternating plus and minus strands are intermediates in adenovirus-associated virus DNA synthesis. Proc. Natl. Acad. Sci. USA. 73: 742–746. | Article | PubMed | ChemPort |
7. Tattersall, P. and Ward, D. C. (1976). Rolling hairpin model for replication of parvovirus and linear chromosomal DNA. Nature. 263: 106–109. | Article | PubMed | ChemPort |
8. Lusby, E., Fife, K. H. and Berns, K. I. (1980). Nucleotide sequence of the inverted terminal repetition in adeno-associated virus DNA. J. Virol. 34: 402–409. | PubMed | ISI | ChemPort |
9. Xiao, X., Xiao, W., Li, J. and Samulski, R. J. (1997). A novel 165-base-pair terminal repeat sequence is the sole cis requirement for the adeno-associated virus life cycle. J. Virol. 71: 941–948. | PubMed | ISI | ChemPort |
10. Berns, K. I. and Linden, R. M. (1995). The cryptic life style of adeno-associated virus. Bioessays. 17: 237–245. | Article | PubMed | ISI | ChemPort |
11. Atchison, R. W., Casto, B. C. and Hammon, W. M. (1965). Adenovirus-associated defective virus particles. Science. 149: 754–756. | Article | PubMed | ISI | ChemPort |
12. Buller, R. M., Janik, J. E., Sebring, E. D. and Rose, J. A. (1981). Herpes simplex virus types 1 and 2 completely help adenovirus-associated virus replication. J. Virol. 40: 241–247. | PubMed | ISI | ChemPort |
13. Muzyczka, N. (1992). Use of adeno-associated virus as a general transduction vector for mammalian cells. Curr. Top. Microbiol. Immunol. 158: 97–129. | PubMed | ISI | ChemPort |
14. Cheung, A. K., Hoggan, M. D., Hauswirth, W. W. and Berns, K. I. (1980). Integration of the adeno-associated virus genome into cellular DNA in latently infected human Detroit 6 cells. J. Virol. 33: 739–748. | PubMed | ISI | ChemPort |
15. Kotin, R. M., et al. (1990). Site-specific integration by adeno-associated virus. Proc. Natl. Acad. Sci. USA. 87: 2211–2215. | Article | PubMed | ChemPort |
16. Xiao, W., et al. (1999). Gene therapy vectors based on adeno-associated virus type 1. J. Virol. 73: 3994–4003. | PubMed | ISI | ChemPort |
17. Rutledge, E. A., Halbert, C. L. and Russell, D. W. (1998). Infectious clones and vectors derived from adeno-associated virus (AAV) serotypes other than AAV type 2. J. Virol. 72: 309–319. | PubMed | ISI | ChemPort |
18. Muramatsu, S., Mizukami, H., Young, N. S. and Brown, K. E. (1996). Nucleotide sequencing and generation of an infectious clone of adeno-associated virus 3. Virology. 221: 208–217. | Article | PubMed | ISI | ChemPort |
19. Gao, G. P., et al. (2002). Novel adeno-associated viruses from rhesus monkeys as vectors for human gene therapy. Proc. Natl. Acad. Sci. USA. 99: 11854–11859. | Article | PubMed | ChemPort |
20. Chiorini, J. A., Kim, F., Yang, L. and Kotin, R. M. (1999). Cloning and characterization of adeno-associated virus type 5. J. Virol. 73: 1309–1319. | PubMed | ISI | ChemPort |
21. Chiorini, J. A., et al. (1997). Cloning of adeno-associated virus type 4 (AAV4) and generation of recombinant AAV4 particles. J. Virol. 71: 6823–6833. | PubMed | ISI | ChemPort |
22. Bantel-Schaal, U., Delius, H., Schmidt, R. and zur Hausen, H. (1999). Human adeno-associated virus type 5 is only distantly related to other known primate helper-dependent parvoviruses. J. Virol. 73: 939–947. | PubMed | ISI | ChemPort |
23. Grimm, D., et al. (2003). Preclinical in vivo evaluation of pseudotyped adeno-associated virus vectors for liver gene therapy. Blood. 102: 2412–2419. | Article | PubMed | ISI | ChemPort |
24. Halbert, C. L., Allen, J. M. and Miller, A. D. (2001). Adeno-associated virus type 6 (AAV6) vectors mediate efficient transduction of airway epithelial cells in mouse lungs compared to that of AAV2 vectors. J. Virol. 75: 6615–6624. | Article | PubMed | ISI | ChemPort |
25. Harper, S. Q., et al. (2002). Modular flexibility of dystrophin: implications for gene therapy of Duchenne muscular dystrophy. Nat. Med. 8: 253–261. | Article | PubMed | ISI | ChemPort |
26. Greelish, J. P., et al. (1999). Stable restoration of the sarcoglycan complex in dystrophic muscle perfused with histamine and a recombinant adeno-associated viral vector. Nat. Med. 5: 439–443. | Article | PubMed | ISI | ChemPort |
27. Blacklow, N. R., et al. (1968). Epidemiology of adenovirus-associated virus infection in a nursery population. Am. J. Epidemiol. 88: 368–378. | PubMed | ChemPort |
28. Grimm, D. and Kleinschmidt, J. A. (1999). Progress in adeno-associated virus type 2 vector production: promises and prospects for clinical use. Hum. Gene Ther. 10: 2445–2450. | Article | PubMed | ISI | ChemPort |
29. Hermonat, P. L., Quirk, J. G., Bishop, B. M. and Han, L. (1997). The packaging capacity of adeno-associated virus (AAV) and the potential for wild-type-plus AAV gene therapy vectors. FEBS Lett. 407: 78–84. | Article | PubMed | ISI | ChemPort |
30. Xiao, X., Li, J. and Samulski, R. J. (1998). Production of high-titer recombinant adeno-associated virus vectors in the absence of helper adenovirus. J. Virol. 72: 2224–2232. | PubMed | ISI | ChemPort |
31. Rabinowitz, J. E., et al. (2002). Cross-packaging of a single adeno-associated virus (AAV) type 2 vector genome into multiple AAV serotypes enables transduction with broad specificity. J. Virol. 76: 791–801. | Article | PubMed | ISI | ChemPort |
32. Chao, H., et al. (2000). Several log increase in therapeutic transgene delivery by distinct adeno-associated viral serotype vectors. Mol. Ther. 2: 619–623. | Article | PubMed | ISI | ChemPort |
33. Herzog, R. W., et al. (1999). Long-term correction of canine hemophilia B by gene transfer of blood coagulation factor IX mediated by adeno-associated viral vector. Nat. Med. 5: 56–63. | Article | PubMed | ISI | ChemPort |
34. Snyder, R. O., et al. (1999). Correction of hemophilia B in canine and murine models using recombinant adeno-associated viral vectors. Nat. Med. 5: 64–70. | Article | PubMed | ISI | ChemPort |
35. Pruchnic, R., et al. (2000). The use of adeno-associated virus to circumvent the maturation-dependent viral transduction of muscle fibers. Hum. Gene Ther. 11: 521–536. | Article | PubMed | ISI | ChemPort |
36. van Deutekom, J. C. and van Ommen, G. J. (2003). Advances in Duchenne muscular dystrophy gene therapy. Nat. Rev. Genet. 4: 774–783. | Article | PubMed | ISI | ChemPort |
37. Mishra, S. K. and Misra, V. (2003). Muscle sarcopenia: an overview. Acta Myol. 22: 43–47. | PubMed | ChemPort |
38. Liu, Y. L., et al. (2003). Optimized production of high-titer recombinant adeno-associated virus in roller bottles. Biotechniques. 34: 184–189. | PubMed | ChemPort |
39. Arruda, V. R., et al. (2004). Safety and efficacy of factor IX gene transfer to skeletal muscle in murine and canine hemophilia B models by adeno-associated viral vector serotype 1. Blood. 103: 85–92. | Article | PubMed | ISI | ChemPort |
40. Liu, F., Nishikawa, M., Clemens, P. R. and Huang, L. (2001). Transfer of full-length Dmd to the diaphragm muscle of Dmd(mdx/mdx) mice through systemic administration of plasmid DNA. Mol. Ther. 4: 45–51. | Article | PubMed | ISI | ChemPort |
41. Mah, C., et al. (2004). A new method for recombinant adeno-associated virus vector delivery to murine diaphragm. Mol. Ther. 9: 458–463. | Article | PubMed | ChemPort |
42. Petrof, B. J., et al. (1995). Efficiency and functional consequences of adenovirus-mediated in vivo gene transfer to normal and dystrophic (mdx) mouse diaphragm. Am. J. Respir. Cell Mol. Biol. 13: 508–517. | PubMed | ChemPort |
43. Huard, J., et al. (1995). The route of administration is a major determinant of the transduction efficiency of rat tissues by adenoviral recombinants. Gene Ther. 2: 107–115. | PubMed | ISI | ChemPort |
44. Huard, J., et al. (1995). Differential short-term transduction efficiency of adult versus newborn mouse tissues by adenoviral recombinants. Exp. Mol. Pathol. 62: 131–143. | Article | PubMed | ChemPort |
45. Gregory, L. G., et al. (2004). Highly efficient EIAV-mediated in utero gene transfer and expression in the major muscle groups affected by Duchenne muscular dystrophy. Gene Ther. 11: 1117–1125. | Article | PubMed | ChemPort |
46. Coutelle, C., et al. (2003). The hopes and fears of in utero gene therapy for genetic disease—a review. Placenta. 24: (Suppl. B): S114–S121. | Article | PubMed | ISI |
47. Wagner, J. A., et al. (2002). A phase II, double-blind, randomized, placebo-controlled clinical trial of tgAAVCF using maxillary sinus delivery in patients with cystic fibrosis with antrostomies. Hum. Gene Ther. 13: 1349–1359. | Article | PubMed | ISI | ChemPort |
48. Moss, R. B., et al. (2004). Repeated adeno-associated virus serotype 2 aerosol-mediated cystic fibrosis transmembrane regulator gene transfer to the lungs of patients with cystic fibrosis: a multicenter, double-blind, placebo-controlled trial. Chest. 125: 509–521. | Article | PubMed | ISI |
49. Kay, M. A., et al. (2000). Evidence for gene transfer and expression of factor IX in haemophilia B patients treated with an AAV vector. Nat. Genet. 24: 257–261. | Article | PubMed | ISI | ChemPort |
50. Gregorevic, P., et al. (2004). Systemic delivery of genes to the striated muscles using adeno-associated viral vectors. Nat. Med. 10: 828–834. (in press). | Article | PubMed | ISI | ChemPort |
51. Schiaffino, S., et al. (1989). Three myosin heavy chain isoforms in type 2 skeletal muscle fibres. J. Muscle Res. Cell Motil. 10: 197–205. | Article | PubMed | ISI | ChemPort |

Acknowledgements

We thank J. A. Kleindschmidt for the pDG packaging plasmid and D. W. D. Russell, G. Miller, and Roli Hirata for the pDGM6 packaging plasmid and helpful discussions. We also thank members of the Chamberlain laboratory for helpful discussions and for critically reading the manuscript. This work was supported by grants from the National Institutes of Health (AG15434 to J.S.C. and DK47754 to A.D.M.) and the Muscular Dystrophy Association (USA) (to J.S.C.).