Gene delivery to the vasculature mediated by low-titre adeno-associated virus serotypes 1 and 5

BACKGROUND

Vascular gene therapy requires safe and efficient gene transfer in vivo. Recombinant adeno-associated virus (AAV) is a promising viral vector but its use in the vasculature has produced conflicting results and serotypes other than AAV2 have not been intensively studied. We investigated the efficiency of alternative AAV serotypes for vascular gene delivery in vitro and in vivo.

METHODS

Vascular cell lines were transduced in vitro with AAV vectors. Rabbit carotid arteries were transduced with AAV1, 2 and 5 encoding enhanced green fluorescent protein (eGFP) ( approximately 1.4 x 10(9) DNAse-resistant particles (drp)). Gene transfer in vivo was assessed at 14 and 28 days. High-titre doses of AAV2 encoding beta-galactosidase in vivo were also studied.

RESULTS

In vitro, transgene expression was not observed in endothelial cells using AAV2 whereas the use of serotypes 1 and 5 resulted in detectable levels of transgene expression. Coronary artery smooth muscle cells (CASMCs) transduced with AAV2 demonstrated higher levels of GFP expression than AAV1 or 5. Transgene expression in vivo was noted using low-titre AAV1 and AAV5 ( approximately 1.4 x 10(9) drp) in the media and adventitia. Only delivery of AAV1eGFP resulted in neointimal formation (3/7 vessels examined), with transgene expression noted in the neointima. Transgene expression with AAV2 was not detected in any layer of the blood vessel wall using low titre ( approximately 10(9) drp). However, high-titre ( approximately 10(11) drp) AAV2 resulted in transduction of cells in the media and adventitia but not the endothelium.

CONCLUSIONS

AAV1 and AAV5 have advantages over AAV2 for vascular gene delivery at low titres.

Mosaic vectors comprised of modified AAV1 capsid proteins for efficient vector purification and targeting to vascular endothelial cells

Vascular-targeted gene therapies have the potential to treat many of the leading causes of mortality in the western world. Unfortunately, these therapies have been ineffective due to poor vascular gene transfer. The use of alternative virus serotypes and the incorporation of vascular targeting ligands into vectors has resulted in only modest increases in vascular gene transfer. Adeno-associated virus (AAV) 1 has shown the most promise among the AAV vectors for the transduction of vascular endothelial cells. However, no straightforward small-scale purification strategy exists for AAV1 as it does for AAV2 making it difficult to quickly produce AAV1 vector for analysis. Here we have combined two AAV1 capsid protein modifications to enhance vascular gene transfer and allow easy purification of vector particles. Mosaic vector particles have been produced comprised of capsid proteins containing the well-characterized RGD4C modification to target integrins present on the vasculature, and capsid proteins containing a modification that permits metabolic biotinylation and efficient purification of mosaic particles by avidin affinity chromatography. We show that the RGD modification results in a 50-100-fold enhancement in endothelial cell gene transfer that is maintained in biotinylated mosaic AAV1 particles. These results suggest that mosaic virions hold significant promise for targeted gene delivery to the vasculature.

Insertional mutagenesis of the adeno-associated virus type 2 (AAV2) capsid gene and generation of AAV2 vectors targeted to alternative cell-surface receptors

Recombinant adeno-associated virus (AAV) vectors are of interest in the context of gene therapy because of their ability to mediate efficient transfer and stable expression of therapeutic genes in a wide variety of tissues. However, AAV-mediated gene delivery to specific cell populations is often precluded by the widespread distribution of heparan sulfate proteoglycan (HSPG), the primary cellular receptor for the virus. Conversely, an increasing number of cell types are being identified that do not express HSPG and are therefore poor targets for AAV-mediated gene transfer. To address these issues, we have developed strategies to physically modify AAV vectors and allow efficient, HSPG-independent, receptor-targeted infection. We began by generating a series of 38 virus capsid mutants containing peptide insertions at 25 unique sites within the AAV capsid protein. The mutant viruses were characterized on the basis of their phenotypes and grouped into three classes: class I mutants (4 of 38) did not assemble particles; class II mutants (14 of 38) assembled noninfectious particles; and class III mutants (20 of 38) assembled fully infectious particles. We examined the HSPG-binding characteristics of the class II mutants and showed that a defect in receptor binding was a common reason for their lack of infectivity. The display of foreign peptide epitopes on the surface of the mutant AAV particles was found to be highly dependent on the inclusion of appropriate scaffolding sequences. Optimal scaffolding sequences and five preferred sites for the insertion of targeting peptide epitopes were identified. These sites are located within each of the three AAV capsid proteins, and thus display inserted epitopes 3, 6, or 60 times per vector particle. Modified AAV vectors displaying a 15-amino acid peptide, which binds to the human luteinizing hormone receptor (LH-R), were generated and assessed for their ability to target gene delivery to receptor-bearing cell lines. Titers of these mutant vectors were essentially the same as wild-type vector. The LH-R-targeted vector was able to transduce ovarian cancer cells (OVCAR-3) in an HSPG-independent manner. Furthermore, transduction was shown to proceed via the LH-R and therefore treatment of OVCAR-3 cells with progesterone, to increase LH-R expression, accordingly increased LH mutant-mediated gene transfer. This technology may have a significant impact on the use of AAV vectors for human gene therapy.

Gene therapy: recombinant adeno-associated virus vectors

Gene transfer using recombinant adeno-associated virus (rAAV) vectors shows great promise for human gene therapy. The broad host range, low level of immune response, and longevity of gene expression observed with these vectors in numerous disease paradigms has enabled the initiation of a number of clinical trials using this gene delivery system. This review presents an overview of the current developments in the field of AAV-mediated gene delivery. Such developments include the establishment of new production methods allowing the generation of high titer preparations, improved purification methods, the use of alternative AAV serotypes, and the generation of trans-splicing rAAV genomes. Together, these developments have improved results interpretation, host range, and the coding capacity of rAAV vectors. Furthermore, the recent identification of regions within the viral capsid that are amenable to modification has begun to address the issue of direct rAAV vector targeting, which could potentially allow targeted gene delivery to specific cell populations. The versatility shown by this vector has enabled new diseases to be realistically considered for therapeutic intervention and considerably broadened the scope of gene therapy.

Infectious entry pathway of adeno-associated virus and adeno-associated virus vectors

We have investigated the infectious entry pathway of adeno-associated virus (AAV) and recombinant AAV vectors by assessing AAV-mediated gene transfer and by covalently conjugating fluorophores to AAV and monitoring entry by fluorescence microscopy. We examined AAV entry in HeLa cells and in HeLa cell lines which inducibly expressed a dominant interfering mutant of dynamin. The data demonstrate that AAV internalizes rapidly by standard receptor-mediated endocytosis from clathrin-coated pits (half-time <10 min). The lysosomotropic agents ammonium chloride and bafilomycin A(1) prevent AAV-mediated gene transfer when present during the first 30 min after the onset of endocytosis, indicating that AAV escapes from early endosomes yet requires an acidic environment for penetration into the cytosol. Following release from the endosome, AAV rapidly moves to the cell nucleus and accumulates perinuclearly beginning within 30 min after the onset of endocytosis. We present data indicating that escape of AAV from the endosome and trafficking of viral particles to the nucleus are unaffected by the presence of adenovirus, the primary helper virus for a productive AAV infection. Within 2 h, viral particles could be detected within the cell nucleus, suggesting that AAV enters the nucleus prior to uncoating. Interestingly, the majority of the intracellular virus particles remain in a stable perinuclear compartment even though gene expression from nuclear AAV genomes can be detected. This suggests that the process of nuclear entry is rate limiting or that AAV entry involves multiple pathways. Nevertheless, these data establish specific points in the AAV infectious entry process and have allowed the generation of a model for future expansion to specific cell types and AAV vector analysis in vivo.

Targeted adeno-associated virus vector transduction of nonpermissive cells mediated by a bispecific F(ab'gamma)2 antibody

We have developed a system for the targeted delivery of adeno-associated virus (AAV) vectors. Targeting is achieved via a bispecific F(ab')2 antibody that mediates a novel interaction between the AAV vector and a specific cell surface receptor expressed on human megakaryocytes. Targeted AAV vectors were able to transduce megakaryocyte cell lines, DAMI and MO7e, which were nonpermissive for normal AAV infection, 70-fold above background and at levels equivalent to permissive K562 cells. Transduction was shown to occur through the specific interaction of the AAV vector-bispecific F(ab')2 complex and cell-associated targeting receptor. Importantly, targeting appeared both selective and restrictive as the endogenous tropism of the AAV vector was significantly reduced. Binding and internalization through the alternative receptor did not alter subsequent steps (escape from endosomes, migration to nucleus, or uncoating) required to successfully transduce target cells. These results demonstrate that AAV vectors can be targeted to a specific cell population and that transduction can be achieved by circumventing the normal virus receptor.

AlphaVbeta5 integrin: a co-receptor for adeno-associated virus type 2 infection

Understanding the primary steps of viral entry can have important implications for strategies to prevent infection of known viral pathogens as well as determining parameters for efficient gene delivery using viral vectors. Recently, a two-step process for viral infection involving attachment of virus to a primary receptor (coxsackievirus adenovirus receptor and heparan sulfate proteoglycan) and subsequent mediation of virus entry by a co-receptor (alphaV integrins and HVEM) has been determined for both adenovirus and HSV, respectively. Heparan sulfate proteoglycan serves as a primary attachment receptor for adeno-associated virus type 2 (AAV-2)(ref. 5). Here we determined that alphaVbeta5 integrin plays a part in efficient AAV infection. Experiments using the chelating agent EDTA to disrupt integrin function resulted in a corresponding decrease in AAV infection, consistent with the possibility that integrin mediates infection. Viral overlay experiments on purified plasma membrane proteins as well as immunoprecipitated integrin beta5 subunit demonstrated that AAV directly associates with the beta5 subunit of alphaVbeta5 integrin. Genetically defined cells expressing alphaVbeta5 integrin showed increased susceptibility to AAV infection, demonstrating a biological role of this integrin in AAV infection. Finally, viral binding and internalization studies indicate that alphaVbeta5 integrin is not a primary attachment receptor for AAV-2, but is instead involved in facilitating virus internalization. This study supports the idea that alphaVbeta5 integrin serves as a co-receptor for AAV-2 virions, and should have a substantial effect on the use of AAV vectors in human gene therapy.

Factors influencing adeno-associated virus-mediated gene transfer to human cystic fibrosis airway epithelial cells: comparison with adenovirus vectors

Adeno-associated virus (AAV) vectors appear promising for use in gene therapy in cystic fibrosis (CF) patients, yet many features of AAV-mediated gene transfer to airway epithelial cells are not well understood. We compared the transduction efficiencies of AAV vectors and adenovirus (Ad) vectors in immortalized cell lines from CF patients and in nasal epithelial primary cultures from normal humans and CF patients. Similar dose-dependent relationships between the vector multiplicities of infection and the efficiencies of lacZ gene transfer were observed. However, levels of transduction for both Ad and recombinant AAV (rAAV) were significantly lower in the airway epithelial cell than in the control cell lines HeLa and HEK 293. Transduction efficiencies differed among cultured epithelial cell types, with poorly differentiated cells transducing more efficiently than well-differentiated cells. A time-dependent increase in gene expression was observed after infection for both vectors. For Ad, but not for AAV, this increase was dependent on prolonged incubation of cells with the vector. Furthermore, for rAAV (but not for rAd), the delay in maximal transduction could be abrogated by wild-type Ad helper infection. Thus, although helper virus is not required for maximal transduction, it increases the kinetics by which this is achieved. Expression of Ad E4 open reading frame 6 or addition of either hydroxyurea or camptothecin resulted in increased AAV transduction, as previously demonstrated for nonairway cells (albeit to lower final levels), suggesting that second-strand synthesis may not be the sole cause of inefficient transduction. Finally, the efficiency of AAV-mediated ex vivo gene transfer to lung cells was similar to that previously described for Ad vectors in that transduction was limited to regions of epithelial injury and preferentially targeted basal-like cells. These studies address the primary factors influencing rAAV infection of human airway cells and should impact successful gene delivery in CF patients.

Raman scattering by pure water and seawater

Measurements of the magnitude and spectral distribution of the Raman-scattering coefficients of pure water (b(rw)) and seawater (b(rs)) are presented. Two independent measurements of the spectral distribution of the Raman-scattering coefficient of pure water were made for incident wavelengths ranging from 250 to 500 nm. These measurements revealed a strong dependence of b(rw) on wavelength that could be represented by a (lambda')(-5.3+/-0.3) relationship, where lambda' is the incident wavelength, or a lambda(-4.6+/-0.3) relationship, where lambda is the Raman-scattered wavelength, when normalized to units of photons. The corresponding relationships for normalization to energy are (lambda')(-5.5+/-0.4) and lambda(-4.8+/-0.3), respectively. These relationships are found to be consistent with resonance Raman theory for an absorption wavelength of 130 nm. The absolute value of b(rw) for the 3400-cm(-1) line was found to be (2.7 +/- 0.2) x 10(-4) m(-1) for an incident wavelength of 488 nm, which is consistent with a number of earlier reports. The difference between the magnitudes of the Raman-scattering coefficients for pure water and seawater was statistically insignificant.

Selective and rapid uptake of adeno-associated virus type 2 in brain

Recombinant adeno-associated virus (AAV) vectors effectively transfer and express foreign genes in the brain. The transferred genes, however, are selectively expressed in neurons, and the cause of this specificity is not understood. To address this question, wild-type AAV-2 capsids were covalently labeled with the fluorophore, Cy3, and infused into the inferior colliculus or the hippocampus. Using antibodies to identify neurons (NeuN), astrocytes (GFAP), or oligodendrocytes (OX-42), clear neuron-specific uptake of the virus was observed as early as 6 min after the start of the infusion. By 30 min postinfusion, AAV particles were present in the nucleus of neurons, yet in both the inferior colliculus and hippocampus, a subset of neurons did not take up the virus particles. No AAV particles were found in astrocytes 1.5 min or 24 hr after virus infusion. Interestingly, 1 hr postinfusion, no AAV particles were found in microglia, yet by 24 hr postinfusion, a punctate pattern of AAV particles was found in microglia. To test whether virus uptake correlated with vector-transduced cells, an rAAV-CMV-GFP virus was infused. By 3 days postinfusion, GFP was localized to neuronal populations with no expression in astrocytes or microglia, similar to that of fluorescent virus uptake. These findings demonstrate that in brain, AAV particles rapidly bind and enter primarily neurons with a pattern similar to that of in vivo vector transduction. In addition, these studies indicate that viral binding and uptake, independent of promoter tropism, can explain the specificity of AAV brain transduction. Thus, this first description of AAV kinetic disposition in vivo should facilitate targeted application of this vector for human brain gene therapy.

Methods for the construction and propagation of recombinant adeno-associated virus vectors

The development of gene transfer vectors from the human parvovirus, adeno-associated virus (AAV), has provided scientists with an efficient and effective way of delivering genes into mammalian cells. This chapter aims to explore the various practical aspects of the AAV vector system, and in consequence, to highlight particular difficulties that may be encountered by workers new to the field. However, before describing the methodology involved in the generation of recombinant AAV vectors, it is of value to briefly discuss the structure and life cycle of this unique virus. Detailed and more extensive reviews that describe the biology of adeno-associated virus are also available (1-3).

Efficient expression of protein coding genes from the murine U1 small nuclear RNA promoters

Few promoters are active at high levels in all cells. Of these, the majority encode structural RNAs transcribed by RNA polymerases I or III and are not accessible for the expression of proteins. An exception are the small nuclear RNAs (snRNAs) transcribed by RNA polymerase II. Although snRNA biosynthesis is unique and thought not to be compatible with synthesis of functional mRNA, we have tested these promoters for their ability to express functional mRNAs. We have used the murine U1a and U1b snRNA gene promoters to express the Escherichia coli lacZ gene and the human alpha-globin gene from either episomal or integrated templates by transfection, or infection into a variety of mammalian cell types. Equivalent expression of beta-galactosidase was obtained from < 250 nucleotides of 5'-flanking sequence containing the complete promoter of either U1 snRNA gene or from the 750-nt cytomegalovirus promoter and enhancer regions. The mRNA was accurately initiated at the U1 start site, efficiently spliced and polyadenylylated, and localized to polyribosomes. Recombinant adenovirus containing the U1b-lacZ chimeric gene transduced and expressed beta-galactosidase efficiently in human 293 cells and airway epithelial cells in culture. Viral vectors containing U1 snRNA promoters may be an attractive alternative to vectors containing viral promoters for persistent high-level expression of therapeutic genes or proteins.

Identification of a tripartite basal promoter which regulates human terminal deoxynucleotidyl transferase gene expression

In order to locate the promoter region of the human terminal deoxynucleotidyl transferase gene, serially truncated segments of the 5'-flanking region of the gene were cloned into a chloramphenicol acetyltransferase reporter vector. Transient transfection analyses of the terminal transferase-reporter gene constructs identified the basal promoter region within -34 to +40 base pairs relative to the transcription start site. Three promoter elements were defined in this region. The primary element is within 34 base pairs upstream of the transcription start site. The CAP site is 62 base pairs upstream of the translation start site. The secondary element involves sequences around the transcription start site. The third is located 25 base pairs downstream from the initiation site (+25 to +40). This tripartite basal promoter was not tissue specific; similar patterns of promoter activity were observed in terminal transferase expressing and non-expressing cells. Transfection analyses also indicated the presence of negative regulatory elements upstream of the basal promoter region, and these elements were preferentially active in cells expressing terminal transferase.