Peter W. Roelvink

CAR-dependent and CAR-independent pathways of adenovirus vector–mediated gene transfer and expression in human fibroblasts

Primary fibroblasts are not efficiently transduced by subgroup C adenovirus (Ad) vectors because they express low levels of the high-affinity Coxsackie virus and adenovirus receptor (CAR). In the present study, we have used primary human dermal fibroblasts as a model to explore strategies by which Ad vectors can be designed to enter cells deficient in CAR. Using an Ad vector expressing the human CAR cDNA (AdCAR) at high multiplicity of infection, primary fibroblasts were converted from being CAR deficient to CAR sufficient. Efficiency of subsequent gene transfer by standard Ad5-based vectors and Ad5-based vectors with alterations in penton and fiber was evaluated. Marked enhancement of binding and transgene expression by standard Ad5 vectors was achieved in CAR-sufficient fibroblasts. Expression by AdDeltaRGDbetagal, an Ad5-based vector lacking the arginine-glycine-aspartate (RGD) alphaV integrin recognition site from its penton base, was achieved in CAR-sufficient, but not CAR-deficient, cells. Fiber-altered Ad5-based vectors, including (a) AdF(pK7)betagal (bearing seven lysines on the end of fiber) (b) AdF(RGD)betagal (bearing a high-affinity RGD sequence on the end of fiber), and (c) AdF9sK betagal (bearing a short fiber and Ad9 knob), demonstrated enhanced gene transfer in CAR-deficient fibroblasts, with no further enhancement in CAR-sufficient fibroblasts. Together, these observations demonstrate that CAR deficiency on Ad targets can be circumvented either by supplying CAR or by modifying the Ad fiber to bind to other cell-surface receptors.

Reducing the Native Tropism of Adenovirus Vectors Requires Removal of both CAR and Integrin Interactions

ABSTRACT The development of tissue-selective virus-based vectors requires a better understanding of the role of receptors in gene transfer in vivo, both to rid the vectors of their native tropism and to introduce new specificity. CAR and αv integrins have been identified as the primary cell surface components that interact with adenovirus type 5 (Ad5)-based vectors during in vitro transduction. We have constructed a set of four vectors, which individually retain the wild-type cell interactions, lack CAR binding, lack αv integrin binding, or lack both CAR and αv integrin binding. These vectors have been used to examine the roles of CAR and αv integrin in determining the tropism of Ad vectors in a mouse model following intrajugular or intramuscular injection. CAR was found to play a significant role in liver transduction. The absence of CAR binding alone, however, had little effect on the low level of expression from Ad in other tissues. Binding of αv integrins appeared to have more influence than did binding of CAR in promoting the expression in these tissues and was also found to be important in liver transduction by Ad vectors. An effect of the penton base modification was a reduction in the number of vector genomes that could be detected in several tissues. In the liver, where CAR binding is important, combining defects in CAR and αv integrin binding was essential to effectively reduce the high level of expression from Ad vectors. While there may be differences in Ad vector tropism among species, our results indicate that both CAR and αv integrins can impact vector distribution in vivo. Disruption of both CAR and αv integrin interactions may be critical for effectively reducing native tropism and enhancing the efficacy of specific targeting ligands in redirecting Ad vectors to target tissues.

Identification of contact residues and definition of the CAR-binding site of adenovirus type 5 fiber protein

ABSTRACT The binding of adenovirus (Ad) fiber knob to its cellular receptor, the coxsackievirus and Ad receptor (CAR), promotes virus attachment to cells and is a major determinant of Ad tropism. Analysis of the kinetics of binding of Ad type 5 (Ad5) fiber knob to the soluble extracellular domains of CAR together (sCAR) and each immunoglobulin (Ig) domain (IgV and IgC2) independently by surface plasmon resonance demonstrated that the IgV domain is necessary and sufficient for binding, and no additional membrane components are required to confer high-affinity binding to Ad5 fiber knob. Four Ad5 fiber knob mutations, Ser408Glu and Pro409Lys in the AB loop, Tyr477Ala in the DG loop, and Leu485Lys in β strand F, effectively abolished high-affinity binding to CAR, while Ala406Lys and Arg412Asp in the AB loop and Arg481Glu in β strand E significantly reduced the level of binding. Circular dichroism spectroscopy showed that these mutations do not disorder the secondary structure of the protein, implicating Ser408, Pro409, Tyr477, and Leu485 as contact residues, with Ala406, Arg412, and Arg481 being peripherally or indirectly involved in CAR binding. The critical residues have exposed side chains that form a patch on the surface, which thus defines the high-affinity interface for CAR. Additional site-directed mutagenesis of Ad5 fiber knob suggests that the binding site does not extend to the adjacent subunit or toward the edge of the R sheet. These findings have implications for our understanding of the biology of Ad infection, the development of novel Ad vectors for targeted gene therapy, and the construction of peptide inhibitors of Ad infection.

Protection against P. aeruginosa with an adenovirus vector containing an OprF epitope in the capsid

Pseudomonas aeruginosa is an important opportunistic pathogen that can cause chronic and often life-threatening infections of the respiratory tract, particularly in individuals with cystic fibrosis (CF). Because infections with P. aeruginosa remain the major cause of the high morbidity and mortality of CF, a vaccine against P. aeruginosa would be very useful for preventing this disorder. The outer membrane protein F (OprF) of P. aeruginosa is a promising vaccine candidate and various B cell epitopes within OprF have been identified. Given that adenovirus (Ad) vectors have strong immunogenic potential and can function as adjuvants for genetic vaccines, the present study evaluates the immunogenic and protective properties of a novel replication-deficient Ad vector in which the Ad hexon protein was modified to include a 14-amino acid epitope of P. aeruginosa OprF (Epi8) in loop 1 of the hypervariable region 5 of the hexon (AdZ.Epi8). Immunization of C57BL/6 mice with AdZ.Epi8 resulted in detectable serum anti-P. aeruginosa and anti-OprF humoral responses. These responses were haplotype dependent, with higher serum anti-OprF titers in CBA mice than in BALB/c or C57BL/6 mice. AdZ.Epi8 induced Epi8-specific IFN-gamma-positive CD4 and CD8 T cell responses and resulted in protection against a lethal pulmonary challenge with agar-encapsulated P. aeruginosa. Importantly, repeated administration of AdZ.Epi8 resulted in boosting of the anti-OprF humoral and anti-Epi8 cellular response, whereas no boosting effect was present in the response against the transgene beta-galactosidase. These observations suggest that Ad vectors expressing pathogen epitopes in their capsid will protect against an extracellular pathogen and will allow boosting of the epitope-specific humoral response with repeated administration, a strategy that should prove useful in developing Ad vectors as vaccines where humoral immunity will be protective.

Efficient and Selective Gene Transfer into Primary Human Brain Tumors by Using Single-Chain Antibody-Targeted Adenoviral Vectors with Native Tropism Abolished

ABSTRACT The application of adenoviral vectors in cancer gene therapy is hampered by low receptor expression on tumor cells and high receptor expression on normal epithelial cells. Targeting adenoviral vectors toward tumor cells may improve cancer gene therapy procedures by providing augmented tumor transduction and decreased toxicity to normal tissues. Targeting requires both the complete abolition of native tropism and the addition of a new specific binding ligand onto the viral capsid. Here we accomplished this by using doubly ablated adenoviral vectors, lacking coxsackievirus-adenovirus receptor and αv integrin binding capacities, together with bispecific single-chain antibodies targeted toward human epidermal growth factor receptor (EGFR) or the epithelial cell adhesion molecule. These vectors efficiently and selectively targeted both alternative receptors on the surface of human cancer cells. Targeted doubly ablated adenoviral vectors were also very efficient and specific with primary human tumor specimens. With primary glioma cell cultures, EGFR targeting augmented the median gene transfer efficiency of doubly ablated adenoviral vectors 123-fold. Moreover, EGFR-targeted doubly ablated vectors were selective for human brain tumors versus the surrounding normal brain tissue. They transduced organotypic glioma and meningioma spheroids with efficiencies similar to those of native adenoviral vectors, while exhibiting greater-than-10-fold-reduced background levels on normal brain explants from the same patients. As a result, EGFR-targeted doubly ablated adenoviral vectors had a 5- to 38-fold-improved tumor-to-normal brain targeting index compared to native vectors. Hence, single-chain targeted doubly ablated adenoviral vectors are promising tools for cancer gene therapy. They should provide an improved therapeutic index with efficient tumor transduction and effective protection of normal tissue.

Epitopes expressed in different adenovirus capsid proteins induce different levels of epitope-specific immunity.

ABSTRACT On the basis of the concept that the capsid proteins of adenovirus (Ad) gene transfer vectors can be genetically manipulated to enhance the immunogenicity of Ad-based vaccines, the present study compared the antiantigen immunogenicity of Ad vectors with a common epitope of the hemagglutinin (HA) protein of the influenza A virus incorporated into the outer Ad capsid protein hexon, penton base, fiber knob, or protein IX. Incorporation of the same epitope into the different capsid proteins provided insights into the correlation between epitope position and antiepitope immunity. Following immunization of three different strains of mice (C57BL/6, BALB/c, and CBA) with either an equal number of Ad particles (resulting in a different total HA copy number) or different Ad particle numbers (to achieve the same HA copy number), the highest primary (immunoglobulin M [IgM]) and secondary (IgG) anti-HA humoral and cellular CD4 gamma interferon and interleukin-4 responses against HA were always achieved with the Ad vector carrying the HA epitope in fiber knob. These observations suggest that the immune response against an epitope inserted into Ad capsid proteins is not necessarily dependent on the capsid protein number and imply that the choice of incorporation site in Ad capsid proteins in their use as vaccines needs to be compared in vivo.

Adenovirus Type 9 Fiber Knob Binds to the Coxsackie B Virus-Adenovirus Receptor (CAR) with Lower Affinity than Fiber Knobs of Other CAR-Binding Adenovirus Serotypes

ABSTRACT The coxsackie B virus and adenovirus (Ad) receptor (CAR) functions as an attachment receptor for multiple Ad serotypes. Here we show that the Ad serotype 9 (Ad9) fiber knob binds to CAR with much reduced affinity compared to the binding by Ad5 and Ad12 fiber knobs as well as the knob of the long fiber of Ad41 (Ad41L). Substitution of Asp222 in Ad9 fiber knob with a lysine that is conserved in Ad5, Ad12, and Ad41L substantially improved Ad9 fiber knob binding to CAR, while the corresponding substitution in Ad5 (Lys442Asp) significantly reduced Ad5 binding. The presence of an aspartic acid residue in Ad9 therefore accounts, at least in part, for the reduced CAR binding affinity of the Ad9 fiber knob. Site-directed mutagenesis of CAR revealed that CAR residues Leu73 and Lys121 and/or Lys123 are critical contact residues, with Tyr80 and Tyr83 being peripherally involved in the binding interaction with the Ad5, Ad9, Ad12, and Ad41L fiber knobs. The overall affinities and the association and dissociation rate constants for wild-type CAR as well as Tyr80 and Tyr83 CAR mutants differed between the serotypes, indicating that their binding modes, although similar, are not identical.

Selective gene transfer into primary human gastric tumors using epithelial cell adhesion molecule-targeted adenoviral vectors with ablated native tropism.

Currently, application of adenoviral vectors (AdV) in gastric cancer gene therapy would be improved by increases in the specificity of transduction. Previously, we found that epithelial cell adhesion molecule (EpCAM) was expressed on gastric tumors but not on gastric epithelium. In this study, we evaluated doubly-ablated AdV lacking native binding ability together with bispecific single-chain antibodies targeted toward EpCAM for gene therapy of gastric cancer. Specific binding to EpCAM augmented the gene transfer efficiency of doubly-ablated AdV on gastric cancer cell lines up to 144-fold, reaching levels similar to or exceeding those achieved with native AdV. In contrast, EpCAM-targeted doubly-ablated AdV-mediated gene transfer into an EpCAM-negative cell line was reduced 38-fold compared with transduction by native AdV. Most importantly, EpCAM-targeted doubly-ablated AdV showed selectivity for primary human gastric tumors versus the surrounding nonneoplastic gastric mucosa of the same patients and normal liver tissue samples. Targeting these doubly-ablated AdV toward EpCAM resulted in similar transduction efficiency as obtained with native AdV for EpCAM-expressing primary human gastric tumors, whereas transduction of gastric epithelium and liver tissue was reduced at least 10-fold. This study thus indicates that application of EpCAM-targeted doubly-ablated AdV for gastric cancer gene therapy results in a favorable tumor-over-normal tissue transduction ratio.

Binding of Adenoviral Fiber Knob to the Coxsackievirus- Adenovirus Receptor Is Crucial for Transduction of Fetal Muscle

Adenoviral (Ad) infection involves attachment mediated by the Ad fiber protein binding to the coxsackievirus-adenovirus receptor (CAR) of a target cell and internalization facilitated by the interaction of the Ad penton base protein with alpha(v) integrins. To understand the relative importance of the Ad binding and internalization steps for the transduction of fetal skeletal muscle, we used a panel of genetically modified vectors that specifically ablate the fiber-CAR interaction (AdL.F*), the penton base-alpha(v) integrin interaction (AdL.PB*), or both (AdL.PB*F*) to transduce embryonic day 16 (E-16) mouse muscle in vivo and primary E-16 muscle cells in vitro. Quantification of transgene expression and vector genome copies revealed a striking absence of E-16 muscle transduction by AdL.F* and AdL.PB*F*. In contrast, fetal muscle transduction with AdL.PB* was not significantly different than with the unmodified vector. Similar results were observed with in vitro Ad infection studies in primary E-16 muscle cells. From these data we conclude that the fiber-CAR interaction is important for the transduction of fetal muscle by Ad vectors. The high dependence on fiber-CAR binding will impact the development of strategies for Ad vector retargeting to achieve muscle-specific transduction in utero.

145. A Comparison of scAAV8-TT-034 Mediated Transduction and shRNA Expression in Human Liver Biopsy Samples versus a Chimeric Mouse Model with Humanized Liver

Background: TT-034 is a DNA-directed RNA interference (ddRNAi) agent designed for the treatment of chronic HCV infection and is currently being tested in a phase I/IIa clinical study. TT-034 is comprised of a vector that expresses three independent short hairpin RNAs (shRNAs) simultaneously targeting three well-conserved regions of the HCV genome. The recombinant genome is packaged in a self-complementary aden-associated virus serotype 8 (AAV8) capsid with tropism for hepatic tissues and delivered as a single dose intravenous (IV) infusion.Methods: Chimeric mouse models in which human hepatocytes replace the majority of mouse hepatocytes are used to study human hepatic function. In order to assess validity of dose/transduction relationships in this murine model to those observed in a human clinical study, chimeric mice were infected with identical doses of TT-034 used in a phase I/IIa study using the same clinical lot of TT-034. In the clinical study, eight subjects have received a single IV infusion of TT-034 at 4.00E10, 1.25E11, 4.00E11 or 1.25E12 vg/kg. At 21 days post dosing, a liver biopsy was collected to assess TT-034 DNA levels and shRNA expression by qPCR. PXB chimeric mice (Phoenix Bio) repopulated with a minimum of 80% hepatocytes were dosed identically with the TT-034 drug product (5 groups, n=4). After 21-28 days, the livers of two mice in each group were removed and hand curated to purify human hepatic tissues (>93% purity). The liver tissues of the other two mice were dissociated and human hepatocytes were enriched to >99% using mouse hepatocyte-capturing Dynabeads. TT-034 DNA and shRNA expression were assessed by qPCR.Results: In the human study, modest levels of TT-034 DNA copies were detected in the 3 subjects dosed at 1.25E11 vg/kg, yielding 0.48, 3.65 and 10.44 copies per cell respectively. Variability in transduction was noted at the higher dose of 4.00E11 vg/kg, with the two subjects yielding 17.74 and 1.01 copies per cell. qPCR analysis of the three anti-HCV shRNAs confirms concomitant, dose dependent expression. In the hand curated samples from the chimeric mouse model (>93% purity), a dose of 1.25E11 vg/kg yielded 1.1 or 1.3 DNA copies per cell while the 4.00E11 vg/kg dose resulted in 1.9 and 5.5 copies DNA per cell. Dynabead-enriched human hepatocytes (>99%) resulted in a considerable drop in DNA copy levels: the 1.25E11 vg/kg dose averaged 0.35 copies per cell while the 4.00E11 vg/kg resulted in 0.85 copies per cell. The lowered DNA levels in the chimeric mouse model led to a concomitant reduction in shRNA expressed from the hand curated tissues. Likewise, shRNA expression was reduced even further in enriched human hepatocytes.Conclusion: Our data suggests that residual mouse hepatocytes present in the chimeric livers are transduced with the scAAV8 vector more efficiently than human hepatocytes and results in lower overall transduction as compared to human clinical samples. Thus, while these models can serve as a surrogate to assess the activity of gene therapy constructs against functions of normal human liver, the doses required for optimal activity may be modestly higher than required in the human clinical setting.