Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2010 Feb 1.
Published in final edited form as: Hear Res. 2008 Dec 7;248(1-2):31–38. doi: 10.1016/j.heares.2008.11.009

Adenoviral vectors for improved gene delivery to the inner ear

Mark Praetorius 1, Douglas E Brough 2, Chi Hsu 2, Peter K Plinkert 1, Hinrich Staecker 3
PMCID: PMC2679534  NIHMSID: NIHMS99479  PMID: 19105978

Abstract

An important requirement for gene therapy in the inner ear is to achieve efficient gene delivery without damaging residual inner ear function. This can be achieved by delivering a high concentration of vector in a minimal volume. Adenovectors are well suited to meet these requirements since high quality concentrated vector with a high capacity for a gene payload can be produced. To reduce the number of vector particles and volume of delivery to the inner ear, we tested vectors with enhancements in cell binding and cell entry properties. We compared delivery of a marker gene to the inner ear using two different advanced generation serotype 5 adenovector designs. The first adenovector tested, AdRGD, has a restricted tropism of entry into cells. AdRGD is an Ad5 capsid vector with an arg-gly-asp (RGD) motif built into the adenovector fiber that has also been modified to abolish the fiber-CAR and penton-integrin interactions that provide the normal well characterized two-step entry pathway for adenovirus. The AdRGD vector has enhanced binding to αV integrins. The second vector, AdF2K, contains 7 lysine residues within the fiber knob and has been shown to have expanded tropism for cells in vitro and in vivo. AdF2K maintains its normal CAR and integrin receptors interactions and has an additional mechanism of entry via its ability to interact with heparan sulfate. Both vectors demonstrated effective delivery to the inner ear and more uniform labeling of the inner ear sensory epithelia than native capsid vector, when tested in vivo. Analysis of expression efficiency using quantitative PCR was tested in vitro on cultured macular organs and demonstrated that vector delivery with the AdF2K vector design yielded optimal delivery. The present study demonstrates that retargeting strategies can improve delivery to the inner ear.

Keywords: Adenovirus vector, enhanced delivery, retargeting, α V integrin, heparan sulfate, gene therapy, inner ear

Introduction

Diseases of the inner ear such as presbycusis, tinnitus and vertigo are widespread, impairing 50 million people in the United States alone. A variety of molecular therapeutic approaches could potentially be developed for these disorders. The inner ear is a good target for gene therapy as it is easily accessed and its fluid spaces are confined, thus enabling high concentrations of vector to be achieved within the inner ear fluid spaces without affecting other organ systems. Many studies have shown efficient gene delivery to structures of the inner ear using a wide variety of vectors including adenovirus, lentivirus (Lalwani, Jero, & Mhatre 2002), herpes simplex virus (Praetorius et al. 2002), adeno-associated virus (Lalwani et al. 1998) and liposomes (Staecker et al. 2001). Adenoviral vectors based on human serotype 5 have been demonstrated as useful for inner ear gene delivery in numerous studies. Many of these studies have demonstrated that despite the delivery of large amounts of vector, delivery of the transgene is not uniform despite using large doses of vector.

Gene delivery using adenovirus to the inner ear is dependant on the presence of native adenovector receptors in the tissue. Entry of the serotype 5 adenovector is mediated by binding of the fiber to the coxsackie-adenovirus-receptor (CAR) which has been shown to be located in tight junctions of cells (Coyne & Bergelson 2005). Therefore it is possible that the CAR is not readily accessible for the vector in many tissues. In the inner ear this may be especially true if the CAR is not presented on the part of the cell accessible via current delivery methods. The second step of this well characterized pathway is internalization of the vector, which is facilitated by binding of αV integrins through an RGD peptide motif on the penton protein of the Ad5 capsid.

Several strategies exist for improving the specificity of gene delivery and ability to interact with target cells. In this study we altered the binding characteristics of the vector thereby changing its tissue/cell type specificity and ability to deliver genes to the inner ear. This approach differs from tissue specific promoter approaches that can increase specificity of transgene expression in selected cells but can not increase the total percentage of delivery to the target cell population. Two different advanced generation adenovector platforms, AdRGD and AdF2K, were examined to enhance delivery into the inner ear. AdRGD has restricted tropism with preferential delivery characteristics to cells expressing αV integrins(Akiyama et al. 2004a;Einfeld et al. 2001b). AdF2K has expanded tropism that binds cells containing CAR, αV integrins, as well as heparan sulfate(Wickham et al. 1996;Wickham et al. 1997). Following injection into the perilymphatic space of the cochlea, both vectors showed more uniform delivery of GFP to the inner ear when compared to native capsid vectors. Analysis of expression efficiency using an adult macular organ culture system, demonstrated that vector delivery with the AdF2K vector design yielded optimal delivery to the inner ear tissue.

Materials and Methods

Animals and Anesthesia

All studies used 3 month-old C57BL/6 mice and were approved by the University of Kansas guidelines for animal care and housing. Anesthesia for surgery was induced using a weight adjusted dose of avertin (0.5 mg/g IP). Post operative pain control was maintained using caprofen (5 mg/kg sq every 12 hrs.).

Vector Production

Three different types of adenovirus vectors were evaluated for gene delivery. 1.) Native Ad5 based adenovectors that expressed either GFP (Adf.11D) or beta-galactosidase (AdZ.11D) were used as controls to evaluate wild type Ad5 capsid interactions for delivery. 2.) A tropism restricted adenovirus vector was evaluated that only interacts with alpha v-integins and 3.) a tropism expanded adenovirus vector was compared that has the ability to use both wt capsid entry and association with heparin sulphate receptors (Adf2K). Construction and production of these adenovirus vector has been previously reported. These adenvirus vectors are E1/E3/E4 deleted and transgene expression is driven by the by the human CMV (hcmv) promoter in each vector. Tropism modified vectors included a vector with an RGD motif in the fiber knob region (AdRGD) and contained modifications that ablated other wild type capsid interactions with CAR and integrins or a vector which carried 7 lysine residues within the fiber (AdF2K) that expands the tropism of the adenovector to additional cell types(Akiyama et al. 2004b;Einfeld et al. 2001a;Wickham, Roelvink, Brough, & Kovesdi 1996;Wickham, Carrion, & Kovesdi 1995). The production system for these adenovectors provided robust replication of the adenovector and purified stocks at 5 × 1011 to 2 × 1012 total particles (particle unit, pu) per ml with a total particle to active particle (fluorescent focus unit, ffu) ratio ranging from 3 – 10 pu/ffu. Total particles (pu) were determined by a spectrophotometric assay that was standardized to reliably and robustly quantify the total particles within a lot of adenovector. Adenovector lots were purified, aliquoted and stored at -80 °C and individual aliquots used for each experiment to prevent loss of activity associated with freeze thaw cycles(Brough et al. 1996;Einfeld et al. 1999).

Vector delivery and histological evaluation

After induction of anesthesia, a dorsal postauricular approach was used to allow exposure of the facial nerve, the posterior semicircular canal and the bulla. A diamond drill was used to open the bulla. After identification of the stapedial artery and round window niche, 1 μl (2 × 107 ffu) of adenovector suspension (Adf.11D, AdZ.11D, AdRGD or AdF2K) was injected into the basal turn of the scala tympani. A small patch of connective tissue was subsequently placed onto the cochleostomy (n=5 mice). A similar procedure was followed for the control mice (n=2) with the exception that 1 μl 0.9 % NaCl solution was injected. After vector delivery animals were allowed to recover and were sacrificed after 48 hours.

Animals were euthanized by intracardiac perfusion with 4% PBS buffered paraformaldehyde. The cochleae were removed and post fixed for two hours in 4% paraformaldehyde in 100mM PBS (pH=7.4) at 4° C. The cochleae were decalcified in Calex (Fisher Diagnostics) for 24 hrs at room temperature. All samples were then dehydrated, cleared and embedded in paraffin. Six μm sections were mounted on sialanated slides (Sigma). Specimens were de-waxed, rehydrated and mounted in anti-fade fluorescent mounting medium. GFP expression was directly observed. Fluorescence was graded by a four point scale from 0 to +++ ranging from absence of fluorescence to saturation. The contra-lateral non injected ear served as a negative control to control for presence of auto-fluorescence. Standard epifluorescence images were photographed with a Nikon 80i microscope with a Magnafire digital camera (Optronics Inc.) using the auto setting.

Immunohistochemistry

Adult C57Bl/6 mice were anaesthetized with intra-peritoneal injection of avertin and sacrificed via intracardiac perfusion with 4% paraformaldehyde. The temporal bones were removed and processed as described above Antibodies used included αV integrin (anti mouse CD51, AbCam Inc.) and heparan sulfate (anti-mouse heparan sulfate, Chemicon Inc.). Secondary staining used a Vectastain ABC (Vector Inc.) kits with a hematoxylin counter-stain performed according to manufacturer directions.

Immunohistochemistry for transgene expression

Expression of beta galactosidase was determined by immunostaining as described above using anti beta galactosidase (Sigma Inc.) and a FITC labelled secondary antibody. GFP expression was determined by immunolabeling with anti-GFP (Abcam Inc.).

Determination of vector transduction efficiency in vitro

Adult C57Bl6 mice were anaesthetized with intraperitoneal avertin and decapitated. The otic capsule was exposed and the macular organs identified by finding the otolithic membranes. Using a #5 watchmaker forceps, the saccule and utricle were removed and the otolith layer dissected away. The organs were cultured in 50 μl of DMEM supplemented with N1 (Sigma) +100 U/ml penicillin and 5.5 μl/ ml of 30% glucose. After 24 hours in vitro (37 °C, 5% CO2) explants were exposed to either 1 × 10 5 ffu or 1 × 10 7 ffu of AdRGD, AdF2K or Adf.11 D for 1 hour. The cultures were then washed 3 times in excess medium and maintained in vitro for 12 hours. Explants were then washed in PBS and then underwent either DNA or RNA extraction. Seven explants were used for each culture condition.

Quantitative PCR

Analysis of Ad genomes:

The primer and probe sequences were as follows: Forward primer: A5s 3825 5′-CGCGGGATTGTGACTGACT -3′. Reverse primer: A5a 3902 5′-GCCAAAAGAGCCGTCAACTT -3′. Fluorogenic Probe OLIGO 1: 5′-FAMAGCAGTGCAGCTTCCCGTTCATCC-TAMRA-3′. A standard curve was generated using eleven serial dilutions of the pAdE1(L)E3(10)E4(WT) plasmid DNA (102 – 106 copies) Quantitative PCR and RT PCR protocols were derived from Applied Biosystems publications cms 042486.pdf and cms 041436.pdf. In addition, a negative control reaction containing no template was assembled. The quantitative PCR reactions were assembled based upon the Taqman PCR Core Reagent kit (PE Applied Biosystems). The reactions were run in duplicate in adjacent wells. Final concentrations of primers and probes were: A5s 3825 (200nM), A5a 3902 (200nM), and OLIGO 1 (100nM). The reactions were thermal cycled using: 50°C for 2 minutes, 95°C for 10 minutes, followed by 40 cycles of 95°C for 15 seconds, and 60°C for 1 minute. Data was collected by the ABI Prism 7700 Sequence Detection System (PE Applied Biosystems) using a standard curve indicating the correlation coefficient and slope. The threshold cycles (Ct) of the no template control reactions were at 40 cycles. The samples were quantified according to their standard deviation and mean in relation to the standard curve.

Quantitative RT-PCR for GFP:

Explants underwent RNA extraction (RNAqueous-4PCR Kit for Isolation of DNA-free RNA, Ambion). One hundred ng of RNA/ reaction were combined added to TaqMan One-Step RT-PCR Master Mix Reagents (Applied Biosystems) along with 100 nM each of GFPs01: 3′-CAT GAG CAA GGG CGA GGA-5′, GFPa62: 3′-CGC CAT CCA GTT CCA CG-5′, and GFP probe : 3′-6FAM CTG TTC ACT GGC GTG GTC CCA ATT C TAMRA-5′. Reverse transcription was carried out 48°C for 30 min, followed by 95°C for 10 min. PCR conditions consisted of 40 cycles of 95°C for 15 seconds, and 60°C for 1 min. For both DNA and RNA quantification, copy number was determined by comparing C(t) results to a standard curve that had been generated. Outcomes were expressed as copies per ng of sample. Statistical analysis: Data was analyzed using 2 way ANOVA using Sigma Stat v 3.5. Significance was set at p≤ 0.05.

Results

Immunohistochemistry to detect αV integrin and heparan sulfate in the inner ear of the mouse

Immunohistochemical labeling for αV integrin was limited to apical region of auditory and vestibular hair cells. There was no labeling of the spiral ganglion, stria vascularis or the spiral ligament (Fig 1). Some labeling was seen at the apical surface of the auditory and vestibular neuroepithelium (1 B and 1D, arrows). Immunohistochemistry for heparan sulfate demonstrated labeling of neurites, mesothelial cells and the stria vascularis. Lower levels of immunolabel are seen in the spiral ligament (Fig 2).

Figure 1.

Figure 1

Open in a new tab

Immunohistochemical labeling (brown) for mouse αV integrin (mouse CD51), the receptor for the AdRGD. All sections are counterstained with hematoxylin resulting in blue staining to the nuclei. An overview of a midmodiolar section at 4x original power is shown in (A). Minimal CD51 (mouse integrin αV) is seen in the apical portion of the vestibular neuroepithelium (B, arrow; 60x original magnification). No significant immunolabelling is seen in the spiral ganglion (C). The organ of Corti demonstrates weak labeling in the inner hair cell (arrow) and along the apex of the outer hair cells (D, 60x original magnification).

Figure 2.

Figure 2

Open in a new tab

Expression of heparan sulfate is seen throughout the inner ear (A, 4x original magnification). There are dense labeling nerve fibers underneath the vestibular neuroepithelium and minimal labeling at the apical portion of the vestibular neuroepithelium (B, 60x original magnification). Expression of heparan sulfate can be seen in the neurites of the spiral ganglion (C, 60x original magnification). Heparan sulfate is also distributed along the basal portion of Corti’s organ but no staining is seen in the hair cells (D, arrows, 60x original magnification)

Distribution of enhanced adenovector gene delivery in vivo

We next looked at the distribution of gene expression delivered from our advanced adenovirus vectors, AdRGD, and AdF2K, as compared to non-capsid-modified control vectors (Adf.11D/AdZ.11D). A low power view of the inner ear after delivery of AdRGD, AdF2K and Adf.11D (Fig. 3) demonstrates distribution of GFP throughout the sensory structures. A section of the contralateral ear from an animal treated with AdRGD is seen in (3D), demonstrating lack of GFP expression on the non-injected side. Both AdRGD and Ad F2K induced robust expression of GFP in the hair cells of the organ of Corti (4 A and B) Adf.11D induced more variable GFP expression with either inner or outer hair cells being transfected (4 C vs. D). The contralateral, non vector treated ear did not demonstrate significant expression of GFP (4E). A second Ad5 capsid control vector (AdZ.11D), transfected the organ of Corti (4F), confirming the efficacy of the Ad5 capsid. All three vectors demonstrated GFP expression in the macular hair cells and supporting cells (Fig 5). Expression of GFP could also be seen in the spiral ganglion (Fig 6). GFP expression characteristics seen with the two modified vectors (AdRGD and AdF2K) and the non-modified control vector (Adf.11D) are summarized in Table 1.

Figure 3.

Figure 3

Open in a new tab

Low power view of the inner ear after delivery of GFP using retargeted adenovectors. GFP fluorescence is seen throughout the inner ear after injection of AdRGD, an integrin αν binding vector, into the scala tympani (A, original magnification 4x). Distribution of GFP after injection of a heparin targeted vector (AdF2K) into the scala tympani is seen in (B). Ad5 capsid vectors (Adf.11D) injected into the scala tympani delivered GFP to the organ of Corti, spiral ganglion and vestibular end organs (C). The contralateral non injected ear from a mouse treated with AdRGD showed only minimal auto-fluorescence in the vestibular ganglion (D).

Figure 5.

Figure 5

Open in a new tab

Transgene expression in the maculae. Expression of GFP could be seen in the macular organs of mice treated with AdRGD (A), AdF2K (B) or Adf.11D (C). Immunofluorescent staining of beta galactosidase in a mouse after delivery of AdZ.11D, demonstrates a transgene expression similar to that seen with GFP expressing vectors (D) (60x original magnification).

Figure 6.

Figure 6

Open in a new tab

Transgene expression in the spiral ganglion: Expression of GFP could be seen in the spiral ganglion of mice treated with AdRGD (A), AdF2K (B) or Adf.11D (C). To rule out auto-fluorescence a subset of sections from animals treated with Adf.11D underwent immunohistochemical staining for GFP (D) demonstrating labeling of spiral ganglion cells (arrows) (60x original magnification).

Table 1.

Expression of GFP in the inner ear for control and modified tropism vectors

Cell type Adf .11D Ad.RGD AdF2K
Stria vascularis (intermediate cells) + + / ++ ++
Inner hair cell -/+ +++ +++
Outer hair cells -/+ + +
Cochlear supporting cells - -/+ -/+
Mesothelial cells (tympanic lamella) - -/+ -/+
Spiral ligament - -/+ -/+
Spiral lamina + + +
Spiral ganglion neurons + + +
Vestibular hair cells + + +
Vestibular supporting cells ++ ++ ++
Open in a new tab

Transfection efficiency for enhanced delivery adenovectors in vitro

In order to evaluate the efficiency of gene delivery we first determined the amount of each vector genome delivered into the inner ear tissue by quantitative PCR. Transfection efficiency was measured by incubating a low (1 × 105 ffu) or high (1 × 107 ffu) dose of active vector particles with adult macular organ cultures for one hour and measuring the vector genomes in the cultured tissue by quantitative PCR. Delivery of a low dose of native capsid vector resulted in the detection of 92227 ± 1598 copies of vector/10 ng DNA sampled, whereas delivery of a high dose of vector resulted in the detection of 781108 ± 11043 copies of vector/10 ng DNA. The amount of AdRGD vector/10 ng DNA sample was 117916 ± 2730 after low and 173090 ± 4598 after high vector dose treatment. After low dose delivery of AdF2K 161920 ± 3829 copies/10 ng DNA were found. Delivery of a high dose AdF2K resulted in 2482975 ± 256331 copies/10 ng DNA (Fig 7). For the high vector doses, delivery of AdRGD was significantly different from Ad.F2K (p<0.001) and significantly different from delivery of the control vector Adf.11D (p<0.001).

Figure 7.

Figure 7

Open in a new tab

The efficiency of vector delivery was evaluated by performing quantitative PCR for vector genomes after exposing adult macular organ cultures to different doses of vector. Delivery of high dose AdF2K achieved the highest level of vector genome delivery in vitro.

Effectiveness of transgene delivery in vitro

In order to determine if one of our advanced adenovector designs yielded enhanced transgene delivery, we next quantitatively determined the level of transgene RNA generated from each vector by quantitative RT-PCR. Adult mouse utricle cultures were treated with either a low (105) or a high (107) number of active vector particles. Subsequently expression of GFP mRNA was measured in each condition using quantitative RT-PCR. The integrin binding vector (AdRGD) showed 49.86 ± 6.8 copies of GFP mRNA/10 ng RNA after low dose vector treatment and 620. 6 ± 33 copies after high dose treatment. The heparan binding vector, AdF2k, demonstrated 830.4 ± 61 copies of GFP mRNA/10ng RNA after low treatment vs. 9109 ± 582 copies after high dose vector treatment. Delivery of native capsid vector (Adf11.D) yielded 60.16 ± 9 copies of GFP mRNA/10ng RNA for low dose treatment and 8665 ± 298 copies GFP mRNA/10 ng RNA for high dose delivery (Fig 8). There was no statistical difference between high dose delivery of control vector and Ad.F2K (p<0.001). Delivery of AdRGD at high doses was significantly different from AdF2K (p<0.001) and Adf.11D (p<0.001). At low doses, delivery of AdF2K resulted in significantly higher production of GFP mRNA compared to AdRGD (p<0.001) or control vector (Adf.11D) (p<0.001).

Figure 8.

Figure 8

Open in a new tab

Transfection efficiency in vitro was determined by exposing adult macular organ cultures to high and low doses of AdRGD, AdF2K and a control vector Adf.11D and measuring expression of GFP mRNA using quantitative RT PCR. High levels of AdF2K and Adf.11D resulted in high levels of delivery of GFP mRNA.

Discussion

In this study we examined the ability of two different adenovector platforms to enhance gene delivery to the inner ear. The main path of entry into the cell for an adenovirus of human serotype 5 (Ad5) is binding of the adenovirus fiber to CAR and entry mediated by interaction with αvβ3 and αvβ5 integrins(Wickham et al. 1993). CAR has been shown to be part of tight junctions and is present at the lateral walls of cells and is present in the inner ear (Venail et al. 2007). Ad5 has also been shown to be able to use other entry pathways. The vascular cell adhesion molecule-1 (Chu et al. 2001), the major histocompatibility complex class I (Hong et al. 1997), the beta2 integrins (Huang et al. 1996), as well as heparan sulfate moieties (Dechecchi et al. 2001) or sialoglycoconjugates (Gaden et al. 2002) have all been demonstrated to facilitate Ad5 entry into cells. The AdRGD vector used in this study is a tropism-restricted vector that does not bind CAR but is directed for entry into cells by RGD engineered into the fiber protein. RGD peptides have been shown to target the alpha-nü integrins on the cell surface. The AdF2K vector on the other hand maintains CAR binding and adds a stretch of lysine residues to the fiber. These poly-lysine residues bind to heparan sulfates at the cell surface (Wickham et al. 1996). We targeted to αV integrins, since integrins of the subtypes alpha nue (V) have been shown to be present in the extra cellular matrix of the inner ear (Tsuprun & Santi 1999), and heparan sulfate, since it is homogeneously expressed in many tissues. Both the integrin and the heparan binding vectors that were used demonstrated transfection of a variety of cell types in vivo (Fig. 3, 4, 5, 6 and Table 1). Based on assays using cultured adult macular organs, the heparan-binding vector appeared more efficient than the integrin-binding vector (Fig. 7 and 8). Taken together these data suggest that modifications to the adenovector fiber/knob region are capable of improving delivery to the inner ear. The distribution of integrin receptors detected by immunohistochemistry did not directly correlate with the distribution of transduction seen with the AdRGD vector. Venail et al. demonstrated a more limited distribution of adenovector-mediated gene expression in the rat cochlea and concluded that distribution of vector corresponded to distribution of CAR and αV integrins(Venail, Wang, Ruel, Ballana, Rebillard, Eybalin, Arbones, Bosch, & Puel 2007). If entry of the AdRGD vector was only through this entry mechanism, then only cells that express integrin should be transduced. Our data does not support this. In the cochlea transfected with the AdRGD, GFP expression can be seen in the modiolus, spiral ganglion cells, Rosenthal’s canal, inner hair cell, some outer hair cells, and in the sensory epithelium of the vestibular organs. Whereas expression of the receptor for the RGD-targeted vector appeared to be seen in a much more limited distribution (Fig. 1). This suggests that this vector may be using alternative intregrin receptors not detected by our immunohistochemistry or that entry mechanisms other than via αV integrin are being used.

Figure 4.

Figure 4

Open in a new tab

Transgene expression in the organ of Corti. AdRGD delivered GFP both to inner and outer hair cells (A, arrows). The expanded tropism vector AdF2K demonstrated a similar distribution of GFP expression (B). Native capsid vector (Adf.11D) yielded variable transfection of hair cells with only outer hair cells (C) or inner hair cells (D) expressing GFP. Controls included the organ of Corti from a contralateral ear (E) and the organ of Corti from a mouse injected with a lac z expressing vector (F). Immunofluorescent staining of beta galactosidase demonstrates a variable pattern of transgene expression similar to that seen with GFP expressing vectors (60x original magnification).

When examining the efficiency of the AdRGD vector using an in vitro culture system, significantly lower levels of vector genome were detected compared to native capsid vector (Adf.11D) or the expanded tropism heparan binding (AdF2K) vector (Fig. 7). However, the amount of GFP expression seen with the AdRGD vector from our in vivo experiments did not correlate to the lower levels of vector genome detected in the in vitro experiments. The contrast between these findings can possibly be explained by the difference in contact time to the cells in each of the experimental systems. In the in vivo experiments, AdRGD remained within the inner ear for the entire experimental period (48 hrs) whereas in the in vitro experiments, any unbound or excess vector was washed out after 1 hour. If the mode of entry in vivo is slower or less efficient via an alternative mechanism then this difference may account for the contrasting finding. Because of ease of manipulations the in vitro experiments offer a system to study the rate of vector entry and may provide a means to further explore the differences in transfection kinetics.

Entry with the heparan binding vector correlated to cells expressing heparan sulfate. In the cochleae transfected with AdF2K, GFP expression can be detected in the modiolus, spiral ganglia cells, stria vascularis, Rosenthal’s canal and to some extent in the inner and outer hair cells. Stronger expression is seen occasionally in vestibular hair cells and at a lower level in the vestibular supporting cells. No overall difference in GFP expression compared to native capsid vector was noted other than that the pattern of GFP expression appeared more uniform. This was unexpected because other tissues such as intestinal mucosa (Lecollinet et al. 2006) and hepatic cells (Schoggins et al. 2005) show larger differences in transfection efficiency of different cell types. Examination of transfection efficiency for the AdF2K vector demonstrated that it performed better than any of the vectors tested when measuring the number of vector genomes present in tissue after in vitro delivery (Fig. 7). When used at low concentrations, AdF2K expressed a higher copy number of GFP mRNA compared to the control vector Adf.11D (Fig. 8). This suggests that this advanced generation vector platform may be useful at lower concentrations than the native capsid vector and may provide enhanced delivery for gene therapy to the inner ear.

Because we did not observe an enhanced number of cells transduced with these new adenovirus vectors we have begun an alternative approach to evaluate vector transduction in the inner ear. The use of quantitative PCR allowed us to determine adenovector genome copy number within the tissue as well as the amount of mRNA expression from the vector. By comparing the amount of expression in utricle cultures to the number of vector genome copies present, we candetermine the efficiency of each retargeted vector. In addition, determination of transgene mRNA levels allows us to determine if modified vectors actually achieve their goal of delivering a gene product. For example, DNA measures of vector genomes will identify vector within cells as well as vector that is bound to cells but has not entered the cell because of changes in the fiber-knob structure. Comparison of the active particle count, the DNA copy number and the transgene mRNA copy number however do give an indication of the efficiency of a particular construct. When combined with a measure of active/inactive particles in a vector batch, we can identify a vectors’ utility in the limited space of the inner ear by evaluating the activity of a batch, knowing how well one vector is delivered versus another and knowing how well an individual vector type is transcribed in a cell. The quantitative (RT-) PCR approach therefore helps us to determine the optimal capsid make up for use in the inner ear. The efficiency of transfection and subsequent gene delivery depends on the vector’s ability to transduce the target cell. The number of vector particles used in a therapeutic vector can have an effect on its beneficial outcomes. The competency of the vector to bind selectively to the target cells is important in terms of reduction of particle numbers needed to transfect efficiently. Modification of the vector’s capsid leads to a change in its ability to direct transduction within the inner ear thereby increasing the specificity and sensitivity of a potential therapeutic. The findings of this paper encourage further research in optimizing the vector-target cell-interaction to optimize targeting and suggest that the modification of vector capsids by cell specific peptides may allow very specific delivery within the inner ear.

Conclusion

Ad5 vector capsids that have undergone modification to increase binding of heparan sulfate increase the transduction efficiency of the vector in the inner ear, especially when low doses of vector were used. Targeting the vector to cellular integrins resulted in a similar histologic distribution of transgene but did not increase transgene delivery. The AdF2K advanced generation adenovirus vector provides an advantage to delivery of transgenes to the inner ear.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Akiyama M, Thorne S, Kirn D, Roelvink PW, Einfeld DA, King CR, Wickham TJ. “Ablating CAR and integrin binding in adenovirus vectors reduces nontarget organ transduction and permits sustained bloodstream persistence following intraperitoneal administration”. Mol.Ther. 2004a;vol. 9(no 2):218–230. doi: 10.1016/j.ymthe.2003.10.010. [DOI] [PubMed] [Google Scholar]
  2. Akiyama M, Thorne S, Kirn D, Roelvink PW, Einfeld DA, King CR, Wickham TJ. “Ablating CAR and integrin binding in adenovirus vectors reduces nontarget organ transduction and permits sustained bloodstream persistence following intraperitoneal administration”. Mol.Ther. 2004b;vol. 9(no 2):218–230. doi: 10.1016/j.ymthe.2003.10.010. [DOI] [PubMed] [Google Scholar]
  3. Brough DE, Lizonova A, Hsu C, Kulesa VA, Kovesdi I. “A gene transfer vector-cell line system for complete functional complementation of adenovirus early regions E1 and E4”. J.Virol. 1996;vol. 70(no 9):6497–6501. doi: 10.1128/jvi.70.9.6497-6501.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chu Y, Heistad D, Cybulsky MI, Davidson BL. “Vascular cell adhesion molecule-1 augments adenovirus-mediated gene transfer”. Arterioscler.Thromb.Vasc.Biol. 2001;vol. 21(no 2):238–242. doi: 10.1161/01.atv.21.2.238. [DOI] [PubMed] [Google Scholar]
  5. Coyne CB, Bergelson JM. “CAR: a virus receptor within the tight junction”. Adv.Drug Deliv.Rev. 2005;vol. 57(no 6):869–882. doi: 10.1016/j.addr.2005.01.007. [DOI] [PubMed] [Google Scholar]
  6. Dechecchi MC, Melotti P, Bonizzato A, Santacatterina M, Chilosi M, Cabrini G. “Heparan sulfate glycosaminoglycans are receptors sufficient to mediate the initial binding of adenovirus types 2 and 5”. J.Virol. 2001;vol. 75(no 18):8772–8780. doi: 10.1128/JVI.75.18.8772-8780.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Einfeld DA, Brough DE, Roelvink PW, Kovesdi I, Wickham TJ. “Construction of a pseudoreceptor that mediates transduction by adenoviruses expressing a ligand in fiber or penton base”. J.Virol. 1999;vol. 73(no 11):9130–9136. doi: 10.1128/jvi.73.11.9130-9136.1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Einfeld DA, Schroeder R, Roelvink PW, Lizonova A, King CR, Kovesdi I, Wickham TJ. “Reducing the native tropism of adenovirus vectors requires removal of both CAR and integrin interactions”. J.Virol. 2001a;vol. 75(no 23):11284–11291. doi: 10.1128/JVI.75.23.11284-11291.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Einfeld DA, Schroeder R, Roelvink PW, Lizonova A, King CR, Kovesdi I, Wickham TJ. “Reducing the native tropism of adenovirus vectors requires removal of both CAR and integrin interactions”. J.Virol. 2001b;vol. 75(no 23):11284–11291. doi: 10.1128/JVI.75.23.11284-11291.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Gaden F, Franqueville L, Hong SS, Legrand V, Figarella C, Boulanger P. “Mechanism of restriction of normal and cystic fibrosis transmembrane conductance regulator-deficient human tracheal gland cells to adenovirus infection and ad-mediated gene transfer”. Am.J.Respir.Cell Mol.Biol. 2002;vol. 27(no 5):628–640. doi: 10.1165/rcmb.4841. [DOI] [PubMed] [Google Scholar]
  11. Hong SS, Karayan L, Tournier J, Curiel DT, Boulanger PA. “Adenovirus type 5 fiber knob binds to MHC class I alpha2 domain at the surface of human epithelial and B lymphoblastoid cells”. EMBO J. 1997;vol. 16(no 9):2294–2306. doi: 10.1093/emboj/16.9.2294. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Huang S, Kamata T, Takada Y, Ruggeri ZM, Nemerow GR. “Adenovirus interaction with distinct integrins mediates separate events in cell entry and gene delivery to hematopoietic cells”. J.Virol. 1996;vol. 70(no 7):4502–4508. doi: 10.1128/jvi.70.7.4502-4508.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Lalwani AK, Jero J, Mhatre AN. “Current issues in cochlear gene transfer”. Audiol.Neurootol. 2002;vol. 7(no 3):146–151. doi: 10.1159/000058300. [DOI] [PubMed] [Google Scholar]
  14. Lalwani AK, Walsh BJ, Carvalho GJ, Muzyczka N, Mhatre AN. “Expression of adeno-associated virus integrated transgene within the mammalian vestibular organs”. Am.J.Otol. 1998;vol. 19(no 3):390–395. [PubMed] [Google Scholar]
  15. Lecollinet S, Gavard F, Havenga MJ, Spiller OB, Lemckert A, Goudsmit J, Eloit M, Richardson J. “Improved gene delivery to intestinal mucosa by adenoviral vectors bearing subgroup B and d fibers”. J.Virol. 2006;vol. 80(no 6):2747–2759. doi: 10.1128/JVI.80.6.2747-2759.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Praetorius M, Knipper M, Schick B, Tan J, Limberger A, Carnicero E, Alonso MT, Schimmang T. “A novel vestibular approach for gene transfer into the inner ear”. Audiol Neurootol. 2002;vol. 7(no 6):324–34. doi: 10.1159/000066157. [DOI] [PubMed] [Google Scholar]
  17. Schoggins JW, Nociari M, Philpott N, Falck-Pedersen E. “Influence of fiber detargeting on adenovirus-mediated innate and adaptive immune activation”. J.Virol. 2005;vol. 79(no 18):11627–11637. doi: 10.1128/JVI.79.18.11627-11637.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Staecker H, Li D, O’Malley BW, Jr., Van de Water TR. “Gene expression in the mammalian cochlea: a study of multiple vector systems”. Acta Otolaryngol. 2001;vol. 121(no 2):157–163. doi: 10.1080/000164801300043307. [DOI] [PubMed] [Google Scholar]
  19. Tsuprun V, Santi P. “Ultrastructure and immunohistochemical identification of the extracellular matrix of the chinchilla cochlea”. Hear.Res. 1999;vol. 129(no 12):35–49. doi: 10.1016/s0378-5955(98)00219-6. [DOI] [PubMed] [Google Scholar]
  20. Venail F, Wang J, Ruel J, Ballana E, Rebillard G, Eybalin M, Arbones M, Bosch A, Puel JL. “Coxsackie adenovirus receptor and alpha nu beta3/alpha nu beta5 integrins in adenovirus gene transfer of rat cochlea”. Gene Ther. 2007;vol. 14(no 1):30–37. doi: 10.1038/sj.gt.3302826. [DOI] [PubMed] [Google Scholar]
  21. Wickham TJ, Carrion ME, Kovesdi I. “Targeting of adenovirus penton base to new receptors through replacement of its RGD motif with other receptor-specific peptide motifs”. Gene Ther. 1995;vol. 2(no 10):750–756. [PubMed] [Google Scholar]
  22. Wickham TJ, Mathias P, Cheresh DA, Nemerow GR. “Integrins alpha v beta 3 and alpha v beta 5 promote adenovirus internalization but not virus attachment”. Cell. 1993;vol. 73(no 2):309–319. doi: 10.1016/0092-8674(93)90231-e. [DOI] [PubMed] [Google Scholar]
  23. Wickham TJ, Roelvink PW, Brough DE, Kovesdi I. “Adenovirus targeted to heparan-containing receptors increases its gene delivery efficiency to multiple cell types”. Nat.Biotechnol. 1996;vol. 14(no 11):1570–1573. doi: 10.1038/nbt1196-1570. [DOI] [PubMed] [Google Scholar]
  24. Wickham TJ, Tzeng E, Shears LL, Roelvink PW, Li Y, Lee GM, Brough DE, Lizonova A, Kovesdi I. “Increased in vitro and in vivo gene transfer by adenovirus vectors containing chimeric fiber proteins”. J.Virol. 1997;vol. 71(no 11):8221–8229. doi: 10.1128/jvi.71.11.8221-8229.1997. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES